Authors: Sapir Caduri
Abstract: Bi-Fact, a novel approach to automatic evaluation for Intent Understanding, is presented. Drawing inspiration from FactScore, Bi-Fact enables fine-grained intent comparison by splitting both gold and predicted intents into facts and calculating precision and recall, considering the UI trajectory. This paper outlines a comprehensive evaluation of Bi-Fact, assessing its performance and comparing it to existing metrics.
Authors: Yuze Zhao, Tianyun Ji, Wenjun Feng, Zhenya Huang, Qi Liu, Zhiding Liu, Yixiao Ma, Kai Zhang, Enhong Chen
Abstract: The reasoning abilities are one of the most enigmatic and captivating aspects of large language models (LLMs). Numerous studies are dedicated to exploring and expanding the boundaries of this reasoning capability. However, tasks that embody both reasoning and recall characteristics are often overlooked. In this paper, we introduce such a novel task, code reasoning, to provide a new perspective for the reasoning abilities of LLMs. We summarize three meta-benchmarks based on established forms of logical reasoning, and instantiate these into eight specific benchmark tasks. Our testing on these benchmarks reveals that LLMs continue to struggle with identifying satisfactory reasoning pathways. Additionally, we present a new pathway exploration pipeline inspired by human intricate problem-solving methods. This Reflective Hypothesis Decomposition and Amendment (RHDA) pipeline consists of the following iterative steps: (1) Proposing potential hypotheses based on observations and decomposing them; (2) Utilizing tools to validate hypotheses and reflection outcomes; (3) Revising hypothesis in light of observations. Our approach effectively mitigates logical chain collapses arising from forgetting or hallucination issues in multi-step reasoning, resulting in performance gains of up to $3\times$. Finally, we expanded this pipeline by applying it to simulate complex household tasks in real-world scenarios, specifically in VirtualHome, enhancing the handling of failure cases. We release our code and all of results at https://github.com/TnTWoW/code_reasoning.
Authors: Alexander Bondarenko, Denis Volk, Dmitrii Volkov, Jeffrey Ladish
Abstract: We demonstrate LLM agent specification gaming by instructing models to win against a chess engine. We find reasoning models like o1 preview and DeepSeek-R1 will often hack the benchmark by default, while language models like GPT-4o and Claude 3.5 Sonnet need to be told that normal play won't work to hack. We improve upon prior work like (Hubinger et al., 2024; Meinke et al., 2024; Weij et al., 2024) by using realistic task prompts and avoiding excess nudging. Our results suggest reasoning models may resort to hacking to solve difficult problems, as observed in OpenAI (2024)'s o1 Docker escape during cyber capabilities testing.
Authors: Soumi Das, Camila Kolling, Mohammad Aflah Khan, Mahsa Amani, Bishwamittra Ghosh, Qinyuan Wu, Till Speicher, Krishna P. Gummadi
Abstract: We study the inherent trade-offs in minimizing privacy risks and maximizing utility, while maintaining high computational efficiency, when fine-tuning large language models (LLMs). A number of recent works in privacy research have attempted to mitigate privacy risks posed by memorizing fine-tuning data by using differentially private training methods (e.g., DP), albeit at a significantly higher computational cost (inefficiency). In parallel, several works in systems research have focussed on developing (parameter) efficient fine-tuning methods (e.g., LoRA), but few works, if any, investigated whether such efficient methods enhance or diminish privacy risks. In this paper, we investigate this gap and arrive at a surprising conclusion: efficient fine-tuning methods like LoRA mitigate privacy risks similar to private fine-tuning methods like DP. Our empirical finding directly contradicts prevailing wisdom that privacy and efficiency objectives are at odds during fine-tuning. Our finding is established by (a) carefully defining measures of privacy and utility that distinguish between memorizing sensitive and non-sensitive tokens in training and test datasets used in fine-tuning and (b) extensive evaluations using multiple open-source language models from Pythia, Gemma, and Llama families and different domain-specific datasets.
Authors: Swati Kar, Soumyabrata Dey, Mahesh K Banavar, Shahnewaz Karim Sakib
Abstract: This paper presents the development of an Artificial Intelligence (AI) based fighter jet agent within a customized Pygame simulation environment, designed to solve multi-objective tasks via deep reinforcement learning (DRL). The jet's primary objectives include efficiently navigating the environment, reaching a target, and selectively engaging or evading an enemy. A reward function balances these goals while optimized hyperparameters enhance learning efficiency. Results show more than 80\% task completion rate, demonstrating effective decision-making. To enhance transparency, the jet's action choices are analyzed by comparing the rewards of the actual chosen action (factual action) with those of alternate actions (counterfactual actions), providing insights into the decision-making rationale. This study illustrates DRL's potential for multi-objective problem-solving with explainable AI. Project page is available at: \href{https://github.com/swatikar95/Autonomous-Fighter-Jet-Navigation-and-Combat}{Project GitHub Link}.
URLs: https://github.com/swatikar95/Autonomous-Fighter-Jet-Navigation-and-Combat
Authors: Xiaojie Xu, Zongyuan Li, Chang Lu, Runnan Qi, Yanan Ni, Lumin Jiang, Xiangbei Liu, Xuebo Zhang, Yongchun Fang, Kuihua Huang, Xian Guo, Zhanghua Wu, Zhenya Li
Abstract: StarCraft II is a complex and dynamic real-time strategy (RTS) game environment, which is very suitable for artificial intelligence and reinforcement learning research. To address the problem of Large Language Model(LLM) learning in complex environments through self-reflection, we propose a Reflection of Episodes(ROE) framework based on expert experience and self-experience. This framework first obtains key information in the game through a keyframe selection method, then makes decisions based on expert experience and self-experience. After a game is completed, it reflects on the previous experience to obtain new self-experience. Finally, in the experiment, our method beat the robot under the Very Hard difficulty in TextStarCraft II. We analyze the data of the LLM in the process of the game in detail, verified its effectiveness.
Authors: Kongcheng Zhang, Qi Yao, Baisheng Lai, Jiaxing Huang, Wenkai Fang, Dacheng Tao, Mingli Song, Shunyu Liu
Abstract: In this work, we propose Reinforced Functional Token Tuning (RFTT), a novel reinforced fine-tuning framework that empowers Large Language Models (LLMs) with self-play learn-to-reason capabilities. Unlike prior prompt-driven reasoning efforts, RFTT embeds a rich set of learnable functional tokens (e.g.,
Authors: Jim Dai, Manxi Wu, Zhanhao Zhang
Abstract: Pioneering companies such as Waymo have deployed robo-taxi services in several U.S. cities. These robo-taxis are electric vehicles, and their operations require the joint optimization of ride matching, vehicle repositioning, and charging scheduling in a stochastic environment. We model the operations of the ride-hailing system with robo-taxis as a discrete-time, average reward Markov Decision Process with infinite horizon. As the fleet size grows, the dispatching is challenging as the set of system state and the fleet dispatching action set grow exponentially with the number of vehicles. To address this, we introduce a scalable deep reinforcement learning algorithm, called Atomic Proximal Policy Optimization (Atomic-PPO), that reduces the action space using atomic action decomposition. We evaluate our algorithm using real-world NYC for-hire vehicle data and we measure the performance using the long-run average reward achieved by the dispatching policy relative to a fluid-based reward upper bound. Our experiments demonstrate the superior performance of our Atomic-PPO compared to benchmarks. Furthermore, we conduct extensive numerical experiments to analyze the efficient allocation of charging facilities and assess the impact of vehicle range and charger speed on fleet performance.
Authors: Hao Ma, Shijie Wang, Zhiqiang Pu, Siyao Zhao, Xiaolin Ai
Abstract: Guiding the policy of multi-agent reinforcement learning to align with human common sense is a difficult problem, largely due to the complexity of modeling common sense as a reward, especially in complex and long-horizon multi-agent tasks. Recent works have shown the effectiveness of reward shaping, such as potential-based rewards, to enhance policy alignment. The existing works, however, primarily rely on experts to design rule-based rewards, which are often labor-intensive and lack a high-level semantic understanding of common sense. To solve this problem, we propose a hierarchical vision-based reward shaping method. At the bottom layer, a visual-language model (VLM) serves as a generic potential function, guiding the policy to align with human common sense through its intrinsic semantic understanding. To help the policy adapts to uncertainty and changes in long-horizon tasks, the top layer features an adaptive skill selection module based on a visual large language model (vLLM). The module uses instructions, video replays, and training records to dynamically select suitable potential function from a pre-designed pool. Besides, our method is theoretically proven to preserve the optimal policy. Extensive experiments conducted in the Google Research Football environment demonstrate that our method not only achieves a higher win rate but also effectively aligns the policy with human common sense.
Authors: Sunder Ali Khowaja, Kapal Dev, Muhammad Salman Pathan, Engin Zeydan, Merouane Debbah
Abstract: We are in a transformative era, and advances in Artificial Intelligence (AI), especially the foundational models, are constantly in the news. AI has been an integral part of many applications that rely on automation for service delivery, and one of them is mission-critical public safety applications. The problem with AI-oriented mission-critical applications is the humanin-the-loop system and the lack of adaptability to dynamic conditions while maintaining situational awareness. Agentic AI (AAI) has gained a lot of attention recently due to its ability to analyze textual data through a contextual lens while quickly adapting to conditions. In this context, this paper proposes an AAI framework for mission-critical applications. We propose a novel framework with a multi-layer architecture to realize the AAI. We also present a detailed implementation of AAI layer that bridges the gap between network infrastructure and missioncritical applications. Our preliminary analysis shows that the AAI reduces initial response time by 5.6 minutes on average, while alert generation time is reduced by 15.6 seconds on average and resource allocation is improved by up to 13.4%. We also show that the AAI methods improve the number of concurrent operations by 40, which reduces the recovery time by up to 5.2 minutes. Finally, we highlight some of the issues and challenges that need to be considered when implementing AAI frameworks.
Authors: Hao Yi, Qingyang Li, Yulan Hu, Fuzheng Zhang, Di Zhang, Yong Liu
Abstract: Recently, enhancing the numerical and logical reasoning capability of Large Language Models (LLMs) has emerged as a research hotspot. Existing methods face several limitations: inference-phase techniques (e.g., Chain of Thoughts) rely on prompt selection and the pretrained knowledge; sentence-level Supervised Fine-Tuning (SFT) and Direct Preference Optimization (DPO) struggle with step-wise mathematical correctness and depend on stronger models distillation or human annotations; while Reinforcement Learning (RL) approaches incur high GPU memory costs and unstable training. To address these, we propose \textbf{S}elf-training framework integrating \textbf{P}rocess \textbf{P}reference learning using \textbf{D}ynamic value margin (SPPD). SPPD leverages a process-based Markov Decision Process (MDP) and Bellman optimality equation to derive \textbf{dynamic value margin} on step-level preference optimization, which employs tree-based self-sampling on model responses \textbf{without any distillation} from other models. Furthermore, we theoretically prove that SPPD is \textbf{equivalent to on-policy policy gradient methods} under reward constraints. Experiments on 7B-scale models demonstrate superior performance across in-domain and out-domain mathematical benchmarks. We open-source our code at \href{https://anonymous.4open.science/r/SSDPO-D-DCDD}{https://anonymous.4open.science/r/SPPD-DCDD}.
URLs: https://anonymous.4open.science/r/SSDPO-D-DCDD, https://anonymous.4open.science/r/SPPD-DCDD
Authors: Yan Yu, Wengang Zhou, Yaodong Yang, Wanxuan Lu, Yingyan Hou, Houqiang Li
Abstract: Multi-task reinforcement learning employs a single policy to complete various tasks, aiming to develop an agent with generalizability across different scenarios. Given the shared characteristics of tasks, the agent's learning efficiency can be enhanced through parameter sharing. Existing approaches typically use a routing network to generate specific routes for each task and reconstruct a set of modules into diverse models to complete multiple tasks simultaneously. However, due to the inherent difference between tasks, it is crucial to allocate resources based on task difficulty, which is constrained by the model's structure. To this end, we propose a Model Evolution framework with Genetic Algorithm (MEGA), which enables the model to evolve during training according to the difficulty of the tasks. When the current model is insufficient for certain tasks, the framework will automatically incorporate additional modules, enhancing the model's capabilities. Moreover, to adapt to our model evolution framework, we introduce a genotype module-level model, using binary sequences as genotype policies for model reconstruction, while leveraging a non-gradient genetic algorithm to optimize these genotype policies. Unlike routing networks with fixed output dimensions, our approach allows for the dynamic adjustment of the genotype policy length, enabling it to accommodate models with a varying number of modules. We conducted experiments on various robotics manipulation tasks in the Meta-World benchmark. Our state-of-the-art performance demonstrated the effectiveness of the MEGA framework. We will release our source code to the public.
Authors: Sylvia S. Kerkhove, Natasha Alechina, Mehdi Dastani
Abstract: Causality plays an important role in daily processes, human reasoning, and artificial intelligence. There has however not been much research on causality in multi-agent strategic settings. In this work, we introduce a systematic way to build a multi-agent system model, represented as a concurrent game structure, for a given structural causal model. In the obtained so-called causal concurrent game structure, transitions correspond to interventions on agent variables of the given causal model. The Halpern and Pearl framework of causality is used to determine the effects of a certain value for an agent variable on other variables. The causal concurrent game structure allows us to analyse and reason about causal effects of agents' strategic decisions. We formally investigate the relation between causal concurrent game structures and the original structural causal models.
Authors: Jessica Lally, Milad Kazemi, Nicola Paoletti
Abstract: This paper addresses a key limitation in existing counterfactual inference methods for Markov Decision Processes (MDPs). Current approaches assume a specific causal model to make counterfactuals identifiable. However, there are usually many causal models that align with the observational and interventional distributions of an MDP, each yielding different counterfactual distributions, so fixing a particular causal model limits the validity (and usefulness) of counterfactual inference. We propose a novel non-parametric approach that computes tight bounds on counterfactual transition probabilities across all compatible causal models. Unlike previous methods that require solving prohibitively large optimisation problems (with variables that grow exponentially in the size of the MDP), our approach provides closed-form expressions for these bounds, making computation highly efficient and scalable for non-trivial MDPs. Once such an interval counterfactual MDP is constructed, our method identifies robust counterfactual policies that optimise the worst-case reward w.r.t. the uncertain interval MDP probabilities. We evaluate our method on various case studies, demonstrating improved robustness over existing methods.
Authors: Hiroyuki Kido
Abstract: An important open question in AI is what simple and natural principle enables a machine to reason logically for meaningful abstraction with grounded symbols. This paper explores a conceptually new approach to combining probabilistic reasoning and predicative symbolic reasoning over data. We return to the era of reasoning with a full joint distribution before the advent of Bayesian networks. We then discuss that a full joint distribution over models of exponential size in propositional logic and of infinite size in predicate logic should be simply derived from a full joint distribution over data of linear size. We show that the same process is not only enough to generalise the logical consequence relation of predicate logic but also to provide a new perspective to rethink well-known limitations such as the undecidability of predicate logic, the symbol grounding problem and the principle of explosion. The reproducibility of this theoretical work is fully demonstrated by the included proofs.
Authors: Juan A. Aledo, Jos\'e A. G\'amez, Alejandro Rosete
Abstract: In rank aggregation problems (RAP), the solution is usually a consensus ranking that generalizes a set of input orderings. There are different variants that differ not only in terms of the type of rankings that are used as input and output, but also in terms of the objective function employed to evaluate the quality of the desired output ranking. In contrast, in some machine learning tasks (e.g. subgroup discovery) or multimodal optimization tasks, attention is devoted to obtaining several models/results to account for the diversity in the input data or across the search landscape. Thus, in this paper we propose to provide, as the solution to an RAP, a set of rankings to better explain the preferences expressed in the input orderings. We exemplify our proposal through the Optimal Bucket Order Problem (OBOP), an RAP which consists in finding a single consensus ranking (with ties) that generalizes a set of input rankings codified as a precedence matrix. To address this, we introduce the Optimal Set of Bucket Orders Problem (OSBOP), a generalization of the OBOP that aims to produce not a single ranking as output but a set of consensus rankings. Experimental results are presented to illustrate this proposal, showing how, by providing a set of consensus rankings, the fitness of the solution significantly improves with respect to the one of the original OBOP, without losing comprehensibility.
Authors: Aleksander Ficek, Somshubra Majumdar, Vahid Noroozi, Boris Ginsburg
Abstract: Code verification has recently found great success as a critical component in training large scale reasoning models for coding. Synthetic techniques such as self-generated test cases and reward models provide a way to enhance code capabilities beyond predefined tests. Building on these advancements, we propose new benchmarks designed to systematically evaluate the impact of synthetic verification methods on assessing solution correctness. We introduce HE-R, HE-R+, MBPP-R, and MBPP-R+, which transform existing coding benchmarks into scoring and ranking datasets to evaluate the effectiveness of synthetic verifiers. Using these benchmarks, we analyze synthetic verification methods in standard, reasoning-based, and reward-based LLMs. Our results show that recent reasoning models significantly improve test case generation and that scaling test cases enhances verification accuracy.
Authors: Zenan Li, Zhaoyu Li, Wen Tang, Xian Zhang, Yuan Yao, Xujie Si, Fan Yang, Kaiyu Yang, Xiaoxing Ma
Abstract: Large language models (LLMs) can prove mathematical theorems formally by generating proof steps (\textit{a.k.a.} tactics) within a proof system. However, the space of possible tactics is vast and complex, while the available training data for formal proofs is limited, posing a significant challenge to LLM-based tactic generation. To address this, we introduce a neuro-symbolic tactic generator that synergizes the mathematical intuition learned by LLMs with domain-specific insights encoded by symbolic methods. The key aspect of this integration is identifying which parts of mathematical reasoning are best suited to LLMs and which to symbolic methods. While the high-level idea of neuro-symbolic integration is broadly applicable to various mathematical problems, in this paper, we focus specifically on Olympiad inequalities (Figure~1). We analyze how humans solve these problems and distill the techniques into two types of tactics: (1) scaling, handled by symbolic methods, and (2) rewriting, handled by LLMs. In addition, we combine symbolic tools with LLMs to prune and rank the proof goals for efficient proof search. We evaluate our framework on 161 challenging inequalities from multiple mathematics competitions, achieving state-of-the-art performance and significantly outperforming existing LLM and symbolic approaches without requiring additional training data.
Authors: Yuliang Liu, Junjie Lu, Zhaoling Chen, Chaofeng Qu, Jason Klein Liu, Chonghan Liu, Zefan Cai, Yunhui Xia, Li Zhao, Jiang Bian, Chuheng Zhang, Wei Shen, Zhouhan Lin
Abstract: Current approaches for training Process Reward Models (PRMs) often involve breaking down responses into multiple reasoning steps using rule-based techniques, such as using predefined placeholder tokens or setting the reasoning step's length into a fixed size. These approaches overlook the fact that specific words do not typically mark true decision points in a text. To address this, we propose AdaptiveStep, a method that divides reasoning steps based on the model's confidence in predicting the next word. This division method provides more decision-making information at each step, enhancing downstream tasks, such as reward model learning. Moreover, our method does not require manual annotation. We demonstrate its effectiveness through experiments with AdaptiveStep-trained PRMs in mathematical reasoning and code generation tasks. Experimental results indicate that the outcome PRM achieves state-of-the-art Best-of-N performance, surpassing greedy search strategy with token-level value-guided decoding, while also reducing construction costs by over 30% compared to existing open-source PRMs. In addition, we provide a thorough analysis and case study on the PRM's performance, transferability, and generalization capabilities.
Authors: Steve Huntsman, Jewell Thomas
Abstract: We devise an algorithm to generate sets of propositions that objectively instantiate graphs that support coherence-driven inference. We then benchmark the ability of large language models (LLMs) to reconstruct coherence graphs from (a straightforward transformation of) propositions expressed in natural language, with promising results from a single prompt to models optimized for reasoning. Combining coherence-driven inference with consistency evaluations by neural models may advance the state of the art in machine cognition.
Authors: Semyon Savkin, Eitan Porat, Or Ordentlich, Yury Polyanskiy
Abstract: Post-training quantization (PTQ) has emerged as a critical technique for efficient deployment of large language models (LLMs). This work proposes NestQuant, a novel PTQ scheme for weights and activations that is based on self-similar nested lattices. Recent work have mathematically shown such quantizers to be information-theoretically optimal for low-precision matrix multiplication. We implement a practical low-complexity version of NestQuant based on Gosset lattice, making it a drop-in quantizer for any matrix multiplication step (e.g., in self-attention, MLP etc). For example, NestQuant quantizes weights, KV-cache, and activations of Llama-3-8B to 4 bits, achieving perplexity of 6.6 on wikitext2. This represents more than 55% reduction in perplexity gap with respect to unquantized model (perplexity of 6.14) compared to state-of-the-art Meta's SpinQuant (perplexity 7.3). Comparisons on various LLM evaluation benchmarks also show a reduction in performance degradation induced by quantization.
Authors: Yiwen Zhang, Yifu Wu, Wenyue Hua, Xuming Hu
Abstract: Large language models have been used to simulate human society using multi-agent systems. Most current social simulation research emphasizes interactive behaviors in fixed environments, ignoring information opacity, relationship variability and diffusion diversity. In this paper, we study the dynamics of information diffusion in 12 asymmetric open environments defined by information content and distribution mechanisms. We first present a general framework to capture the features of information diffusion. Then, we designed a dynamic attention mechanism to help agents allocate attention to different information, addressing the limitations of LLM-based attention. Agents start by responding to external information stimuli within a five-agent group, increasing group size and forming information circles while developing relationships and sharing information. Additionally, we observe the emergence of information cocoons, the evolution of information gaps, and the accumulation of social capital, which are closely linked to psychological, sociological, and communication theories.
Authors: Maxwell J. D. Ramstead, Candice Pattisapu, Jason Fox, Jeff Beck
Abstract: This white paper describes some of the design principles for artificial or machine intelligence that guide efforts at Noumenal Labs. These principles are drawn from both nature and from the means by which we come to represent and understand it. The end goal of research and development in this field should be to design machine intelligences that augment our understanding of the world and enhance our ability to act in it, without replacing us. In the first two sections, we examine the core motivation for our approach: resolving the grounding problem. We argue that the solution to the grounding problem rests in the design of models grounded in the world that we inhabit, not mere word models. A machine super intelligence that is capable of significantly enhancing our understanding of the human world must represent the world as we do and be capable of generating new knowledge, building on what we already know. In other words, it must be properly grounded and explicitly designed for rational, empirical inquiry, modeled after the scientific method. A primary implication of this design principle is that agents must be capable of engaging autonomously in causal physics discovery. We discuss the pragmatic implications of this approach, and in particular, the use cases in realistic 3D world modeling and multimodal, multidimensional time series analysis.
Authors: Ziyi Ni, Hao Wang, Huacan Wang
Abstract: Large Language Models (LLMs) have achieved remarkable success in various domains but remain vulnerable to adversarial jailbreak attacks. Existing prompt-defense strategies, including parameter-modifying and parameter-free approaches, face limitations in adaptability, interpretability, and customization, constraining their effectiveness against evolving threats. To address these challenges, we propose ShieldLearner, a novel paradigm that mimics human learning in defense. Through trial and error, it autonomously distills attack signatures into a Pattern Atlas and synthesizes defense heuristics into a Meta-analysis Framework, enabling systematic and interpretable threat detection. Furthermore, we introduce Adaptive Adversarial Augmentation to generate adversarial variations of successfully defended prompts, enabling continuous self-improvement without model retraining. In addition to standard benchmarks, we create a hard test set by curating adversarial prompts from the Wildjailbreak dataset, emphasizing more concealed malicious intent. Experimental results show that ShieldLearner achieves a significantly higher defense success rate than existing baselines on both conventional and hard test sets, while also operating with lower computational overhead, making it a practical and efficient solution for real-world adversarial defense.
Authors: Mohammad Wali Ur Rahman, Ric Nevarez, Lamia Tasnim Mim, Salim Hariri
Abstract: In this paper, we introduce MASQRAD (Multi-Agent Strategic Query Resolution and Diagnostic tool), a transformative framework for query resolution based on the actor-critic model, which utilizes multiple generative AI agents. MASQRAD is excellent at translating imprecise or ambiguous user inquiries into precise and actionable requests. This framework generates pertinent visualizations and responses to these focused queries, as well as thorough analyses and insightful interpretations for users. MASQRAD addresses the common shortcomings of existing solutions in domains that demand fast and precise data interpretation, such as their incapacity to successfully apply AI for generating actionable insights and their challenges with the inherent ambiguity of user queries. MASQRAD functions as a sophisticated multi-agent system but "masquerades" to users as a single AI entity, which lowers errors and enhances data interaction. This approach makes use of three primary AI agents: Actor Generative AI, Critic Generative AI, and Expert Analysis Generative AI. Each is crucial for creating, enhancing, and evaluating data interactions. The Actor AI generates Python scripts to generate data visualizations from large datasets within operational constraints, and the Critic AI rigorously refines these scripts through multi-agent debate. Finally, the Expert Analysis AI contextualizes the outcomes to aid in decision-making. With an accuracy rate of 87\% when handling tasks related to natural language visualization, MASQRAD establishes new benchmarks for automated data interpretation and showcases a noteworthy advancement that has the potential to revolutionize AI-driven applications.
Authors: Xiangyu Li, Yawen Zeng, Xiaofen Xing, Jin Xu, Xiangmin Xu
Abstract: As automated trading gains traction in the financial market, algorithmic investment strategies are increasingly prominent. While Large Language Models (LLMs) and Agent-based models exhibit promising potential in real-time market analysis and trading decisions, they still experience a significant -20% loss when confronted with rapid declines or frequent fluctuations, impeding their practical application. Hence, there is an imperative to explore a more robust and resilient framework. This paper introduces an innovative multi-agent system, HedgeAgents, aimed at bolstering system robustness via ``hedging'' strategies. In this well-balanced system, an array of hedging agents has been tailored, where HedgeAgents consist of a central fund manager and multiple hedging experts specializing in various financial asset classes. These agents leverage LLMs' cognitive capabilities to make decisions and coordinate through three types of conferences. Benefiting from the powerful understanding of LLMs, our HedgeAgents attained a 70% annualized return and a 400% total return over a period of 3 years. Moreover, we have observed with delight that HedgeAgents can even formulate investment experience comparable to those of human experts (https://hedgeagents.github.io/).
Authors: Jun Zhuang, Chaowen Guan
Abstract: In the era of noisy intermediate-scale quantum (NISQ) computing, Quantum Neural Networks (QNNs) have emerged as a promising approach for various applications, yet their training is often hindered by barren plateaus (BPs), where gradient variance vanishes exponentially as the model size increases. To address this challenge, we propose a new Large Language Model (LLM)-driven search framework, AdaInit, that iteratively searches for optimal initial parameters of QNNs to maximize gradient variance and therefore mitigate BPs. Unlike conventional one-time initialization methods, AdaInit dynamically refines QNN's initialization using LLMs with adaptive prompting. Theoretical analysis of the Expected Improvement (EI) proves a supremum for the search, ensuring this process can eventually identify the optimal initial parameter of the QNN. Extensive experiments across four public datasets demonstrate that AdaInit significantly enhances QNN's trainability compared to classic initialization methods, validating its effectiveness in mitigating BPs.
Authors: Jun Kevin, Pujianto Yugopuspito
Abstract: Smart contracts are essential to decentralized finance (DeFi) and blockchain ecosystems but are increasingly vulnerable to exploits due to coding errors and complex attack vectors. Traditional static analysis tools and existing vulnerability detection methods often fail to address these challenges comprehensively, leading to high false-positive rates and an inability to detect dynamic vulnerabilities. This paper introduces SmartLLM, a novel approach leveraging fine-tuned LLaMA 3.1 models with Retrieval-Augmented Generation (RAG) to enhance the accuracy and efficiency of smart contract auditing. By integrating domain-specific knowledge from ERC standards and employing advanced techniques such as QLoRA for efficient fine-tuning, SmartLLM achieves superior performance compared to static analysis tools like Mythril and Slither, as well as zero-shot large language model (LLM) prompting methods such as GPT-3.5 and GPT-4. Experimental results demonstrate a perfect recall of 100% and an accuracy score of 70%, highlighting the model's robustness in identifying vulnerabilities, including reentrancy and access control issues. This research advances smart contract security by offering a scalable and effective auditing solution, supporting the secure adoption of decentralized applications.
Authors: Muhammad Fahad Zia, Sri Harish Kalidass
Abstract: Phishing is the most prevalent type of cyber-attack today and is recognized as the leading source of data breaches with significant consequences for both individuals and corporations. Web-based phishing attacks are the most frequent with vectors such as social media posts and emails containing links to phishing URLs that once clicked on render host systems vulnerable to more sinister attacks. Research efforts to detect phishing URLs have involved the use of supervised learning techniques that use large amounts of data to train models and have high computational requirements. They also involve analysis of features derived from vectors including email contents thus affecting user privacy. Additionally, they suffer from a lack of resilience against evolution of threats especially with the advent of generative AI techniques to bypass these systems as with AI-generated phishing URLs. Unsupervised methods such as clustering techniques have also been used in phishing detection in the past, however, they are at times unscalable due to the use of pair-wise comparisons. They also lack high detection rates while detecting phishing campaigns. In this paper, we propose an unsupervised learning approach that is not only fast but scalable, as it does not involve pair-wise comparisons. It is able to detect entire campaigns at a time with a high detection rate while preserving user privacy; this includes the recent surge of campaigns with targeted phishing URLs generated by malicious entities using generative AI techniques.
Authors: Bo Wang, Weiyi He, Pengfei He, Shenglai Zeng, Zhen Xiang, Yue Xing, Jiliang Tang
Abstract: Large Language Model (LLM) agents have become increasingly prevalent across various real-world applications. They enhance decision-making by storing private user-agent interactions in the memory module for demonstrations, introducing new privacy risks for LLM agents. In this work, we systematically investigate the vulnerability of LLM agents to our proposed Memory EXTRaction Attack (MEXTRA) under a black-box setting. To extract private information from memory, we propose an effective attacking prompt design and an automated prompt generation method based on different levels of knowledge about the LLM agent. Experiments on two representative agents demonstrate the effectiveness of MEXTRA. Moreover, we explore key factors influencing memory leakage from both the agent's and the attacker's perspectives. Our findings highlight the urgent need for effective memory safeguards in LLM agent design and deployment.
Authors: Wang Yang, Hongye Jin, Jingfeng Yang, Vipin Chaudhary, Xiaotian Han
Abstract: Supervised Fine-Tuning (SFT) has been a go-to and effective method for enhancing long chain-of-thought (CoT) reasoning in relatively small LLMs by fine-tuning them with long CoT responses from larger LLMs. To continually improve reasoning abilities, we can either collect new high-quality long CoT reasoning SFT data or repeatedly train on existing SFT datasets. However, acquiring new long CoT SFT data is costly and limited, while repeated training often results in a performance plateau or decline. To further boost the performance with the SFT data, we propose Thinking Preference Optimization (ThinkPO), a simple yet effective post-SFT method that enhances long CoT reasoning without requiring new long CoT responses. Instead, ThinkPO utilizes readily available or easily obtainable short CoT reasoning responses as rejected answers and long CoT responses as chosen answers for the same question. It then applies direct preference optimization to encourage the model to favor longer reasoning outputs. Experiments show that ThinkPO further improves the reasoning performance of SFT-ed models, e.g. it increases math reasoning accuracy of SFT-ed models by 8.6% and output length by 25.9%. Notably, ThinkPO is capable of continually boosting the performance of the publicly distilled SFT model, e.g., increasing the official DeepSeek-R1-Distill-Qwen-7B's performance on MATH500 from 87.4% to 91.2%.
Authors: Andreas Radler, Eric Volkmann, Johannes Brandstetter, Arturs Berzins
Abstract: Topology optimization (TO) is a family of computational methods that derive near-optimal geometries from formal problem descriptions. Despite their success, established TO methods are limited to generating single solutions, restricting the exploration of alternative designs. To address this limitation, we introduce Generative Topology Optimization (GenTO) - a data-free method that trains a neural network to generate structurally compliant shapes and explores diverse solutions through an explicit diversity constraint. The network is trained with a solver-in-the-loop, optimizing the material distribution in each iteration. The trained model produces diverse shapes that closely adhere to the design requirements. We validate GenTO on 2D and 3D TO problems. Our results demonstrate that GenTO produces more diverse solutions than any prior method while maintaining near-optimality and being an order of magnitude faster due to inherent parallelism. These findings open new avenues for engineering and design, offering enhanced flexibility and innovation in structural optimization.
Authors: Wenpeng Xing, Minghao Li, Mohan Li, Meng Han
Abstract: Embodied AI systems, including robots and autonomous vehicles, are increasingly integrated into real-world applications, where they encounter a range of vulnerabilities stemming from both environmental and system-level factors. These vulnerabilities manifest through sensor spoofing, adversarial attacks, and failures in task and motion planning, posing significant challenges to robustness and safety. Despite the growing body of research, existing reviews rarely focus specifically on the unique safety and security challenges of embodied AI systems. Most prior work either addresses general AI vulnerabilities or focuses on isolated aspects, lacking a dedicated and unified framework tailored to embodied AI. This survey fills this critical gap by: (1) categorizing vulnerabilities specific to embodied AI into exogenous (e.g., physical attacks, cybersecurity threats) and endogenous (e.g., sensor failures, software flaws) origins; (2) systematically analyzing adversarial attack paradigms unique to embodied AI, with a focus on their impact on perception, decision-making, and embodied interaction; (3) investigating attack vectors targeting large vision-language models (LVLMs) and large language models (LLMs) within embodied systems, such as jailbreak attacks and instruction misinterpretation; (4) evaluating robustness challenges in algorithms for embodied perception, decision-making, and task planning; and (5) proposing targeted strategies to enhance the safety and reliability of embodied AI systems. By integrating these dimensions, we provide a comprehensive framework for understanding the interplay between vulnerabilities and safety in embodied AI.
Authors: Ahmed Burak Gulhan, Krishna Teja Chitty-Venkata, Murali Emani, Mahmut Kandemir, Venkatram Vishwanath
Abstract: In Large Language Model (LLM) inference, Key-Value (KV) caches (KV-caches) are essential for reducing time complexity. However, they result in a linear increase in GPU memory as the context length grows. While recent work explores KV-cache eviction and compression policies to reduce memory usage, they often consider uniform KV-caches across all attention heads, leading to suboptimal performance. We introduce BaKlaVa, a method to allocate optimal memory for individual KV-caches across the model by estimating the importance of each KV-cache. Our empirical analysis demonstrates that not all KV-caches are equally critical for LLM performance. Using a one-time profiling approach, BaKlaVa assigns optimal memory budgets to each KV-cache. We evaluated our method on LLaMA-3-8B, and Qwen2.5-7B models, achieving up to a 70\% compression ratio while keeping baseline performance and delivering up to an order-of-magnitude accuracy improvement at higher compression levels.
Authors: Sangkyu Lee, Janghoon Han, Hosung Song, Stanley Jungkyu Choi, Honglak Lee, Youngjae Yu
Abstract: Direct Preference Optimization (DPO) demonstrates the advantage of aligning a large language model with human preference using only an offline dataset. However, DPO has the limitation that the KL penalty, which prevents excessive deviation from the reference model, is static throughout the training process. Several methods try to turn this static KL penalty into a dynamic one, but no approach can adaptively assign different KL penalties for each preference pair. In this paper, we propose $\varepsilon$-Direct Preference Optimization ($\varepsilon$-DPO), which allows adaptive control of the KL penalty strength $\beta$ for each preference pair. Specifically, $\varepsilon$-DPO adaptively controls $\beta$ for each preference pair based on the monotonicity of logits as a preference model under the perturbation of $\beta$ during training by simply reusing the logit of the current policy and the reference policy. Experimental results show that $\varepsilon$-DPO outperforms existing direct alignment algorithms and KL penalty relaxation methods on general chatbot benchmarks, highlighting the significance of adaptive KL penalty relaxation at the instance-level in DPO.
Authors: Jiaqi Zhao, Ming Wang, Miao Zhang, Yuzhang Shang, Xuebo Liu, Yaowei Wang, Min Zhang, Liqiang Nie
Abstract: Post-training Quantization (PTQ) technique has been extensively adopted for large language models (LLMs) compression owing to its efficiency and low resource requirement. However, current research lacks a in-depth analysis of the superior and applicable scenarios of each PTQ strategy. In addition, existing algorithms focus primarily on performance, overlooking the trade-off among model size, performance, and quantization bitwidth. To mitigate these confusions, we provide a novel benchmark for LLMs PTQ in this paper. Firstly, in order to support our benchmark, we propose a comprehensive taxonomy for existing mainstream methods by scrutinizing their computational strategies (e.g., optimization-based, compensation-based, etc.). Then, we conduct extensive experiments with the baseline within each class, covering models with various sizes (7B-70B), bitwidths, training levels (LLaMA1/2/3/3.1), architectures (Mixtral, DeepSeekMoE and Mamba) and modality (LLaVA1.5 and VILA1.5) on a wide range of evaluation metrics.Through comparative analysis on the results, we summarize the superior of each PTQ strategy and modelsize-bitwidth trade-off considering the performance. For example, our benchmark reveals that compensation-based technique demonstrates outstanding cross-architecture robustness and extremely low-bit PTQ for ultra large models should be reexamined. Finally, we further accordingly claim that a practical combination of compensation and other PTQ strategy can achieve SOTA various robustness. We believe that our benchmark will provide valuable recommendations for the deployment of LLMs and future research on PTQ approaches.
Authors: Jiaqi Zhao, Miao Zhang, Ming Wang, Yuzhang Shang, Kaihao Zhang, Weili Guan, Yaowei Wang, Min Zhang
Abstract: Large Language Models (LLMs) suffer severe performance degradation when facing extremely low-bit (sub 2-bit) quantization. Several existing sub 2-bit post-training quantization (PTQ) methods utilize a mix-precision scheme by leveraging an unstructured fine-grained mask to explicitly distinguish salient weights, while which introduces an extra 1-bit or more per weight. To explore the real limit of PTQ, we propose an extremely low-bit PTQ method called PTQ1.61, which enables weight quantization to 1.61-bit for the first time. Specifically, we first introduce a one-dimensional structured mask with negligibly additional 0.0002-bit per weight based on input activations from the perspective of reducing the upper bound of quantization error to allocate corresponding salient weight channels to 4-bit. For non-salient channels binarization, an efficient block-wise scaling factors optimization framework is then presented to take implicit row-wise correlations and angular biases into account. Different from prior works that concentrate on adjusting quantization methodologies, we further propose a novel paradigm called quantization preprocessing, where we argue that transforming the weight distribution of the pretrained model before quantization can alleviate the difficulty in per-channel extremely low-bit PTQ. Extensive experiments indicate our PTQ1.61 achieves state-of-the-art performance in extremely low-bit quantization. Codes are available at https://github.com/zjq0455/PTQ1.61.
Authors: Hongxu Wang, Zhu Sun, Yingpeng Du, Lu Zhang, Tiantian He, Yew-Soon Ong
Abstract: Recommender systems (RSs) play a crucial role in shaping our digital interactions, influencing how we access and engage with information across various domains. Traditional research has predominantly centered on maximizing recommendation accuracy, often leading to unintended side effects such as echo chambers and constrained user experiences. Drawing inspiration from autonomous driving, we introduce a novel framework that categorizes RS autonomy into five distinct levels, ranging from basic rule-based accuracy-driven systems to behavior-aware, uncertain multi-objective RSs - where users may have varying needs, such as accuracy, diversity, and fairness. In response, we propose an approach that dynamically identifies and optimizes multiple objectives based on individual user preferences, fostering more ethical and intelligent user-centric recommendations. To navigate the uncertainty inherent in multi-objective RSs, we develop a Bayesian optimization (BO) framework that captures personalized trade-offs between different objectives while accounting for their uncertain interdependencies. Furthermore, we introduce an orthogonal meta-learning paradigm to enhance BO efficiency and effectiveness by leveraging shared knowledge across similar tasks and mitigating conflicts among objectives through the discovery of orthogonal information. Finally, extensive empirical evaluations demonstrate the effectiveness of our method in optimizing uncertain multi-objectives for individual users, paving the way for more adaptive and user-focused RSs.
Authors: Jaemu Heo, Eldor Fozilov, Hyunmin Song, Taehwan Kim
Abstract: Transformers have achieved great success in effectively processing sequential data such as text. Their architecture consisting of several attention and feedforward blocks can model relations between elements of a sequence in parallel manner, which makes them very efficient to train and effective in sequence modeling. Even though they have shown strong performance in processing sequential data, the size of their parameters is considerably larger when compared to other architectures such as RNN and CNN based models. Therefore, several approaches have explored parameter sharing and recurrence in Transformer models to address their computational demands. However, such methods struggle to maintain high performance compared to the original transformer model. To address this challenge, we propose our novel approach, RingFormer, which employs one Transformer layer that processes input repeatedly in a circular, ring-like manner, while utilizing low-rank matrices to generate input-dependent level signals. This allows us to reduce the model parameters substantially while maintaining high performance in a variety of tasks such as translation and image classification, as validated in the experiments.
Authors: Xin Wang, Juntao Yang, Jeff Adie, Simon See, Kalli Furtado, Chen Chen, Troy Arcomano, Romit Maulik, Gianmarco Mengaldo
Abstract: Accurate and efficient climate simulations are crucial for understanding Earth's evolving climate. However, current general circulation models (GCMs) face challenges in capturing unresolved physical processes, such as cloud and convection. A common solution is to adopt cloud resolving models, that provide more accurate results than the standard subgrid parametrisation schemes typically used in GCMs. However, cloud resolving models, also referred to as super paramtetrizations, remain computationally prohibitive. Hybrid modeling, which integrates deep learning with equation-based GCMs, offers a promising alternative but often struggles with long-term stability and accuracy issues. In this work, we find that water vapor oversaturation during condensation is a key factor compromising the stability of hybrid models. To address this, we introduce CondensNet, a novel neural network architecture that embeds a self-adaptive physical constraint to correct unphysical condensation processes. CondensNet effectively mitigates water vapor oversaturation, enhancing simulation stability while maintaining accuracy and improving computational efficiency compared to super parameterization schemes. We integrate CondensNet into a GCM to form PCNN-GCM (Physics-Constrained Neural Network GCM), a hybrid deep learning framework designed for long-term stable climate simulations in real-world conditions, including ocean and land. PCNN-GCM represents a significant milestone in hybrid climate modeling, as it shows a novel way to incorporate physical constraints adaptively, paving the way for accurate, lightweight, and stable long-term climate simulations.
Authors: Longchao Da, Justin Turnau, Thirulogasankar Pranav Kutralingam, Alvaro Velasquez, Paulo Shakarian, Hua Wei
Abstract: Deep Reinforcement Learning (RL) has been explored and verified to be effective in solving decision-making tasks in various domains, such as robotics, transportation, recommender systems, etc. It learns from the interaction with environments and updates the policy using the collected experience. However, due to the limited real-world data and unbearable consequences of taking detrimental actions, the learning of RL policy is mainly restricted within the simulators. This practice guarantees safety in learning but introduces an inevitable sim-to-real gap in terms of deployment, thus causing degraded performance and risks in execution. There are attempts to solve the sim-to-real problems from different domains with various techniques, especially in the era with emerging techniques such as large foundations or language models that have cast light on the sim-to-real. This survey paper, to the best of our knowledge, is the first taxonomy that formally frames the sim-to-real techniques from key elements of the Markov Decision Process (State, Action, Transition, and Reward). Based on the framework, we cover comprehensive literature from the classic to the most advanced methods including the sim-to-real techniques empowered by foundation models, and we also discuss the specialties that are worth attention in different domains of sim-to-real problems. Then we summarize the formal evaluation process of sim-to-real performance with accessible code or benchmarks. The challenges and opportunities are also presented to encourage future exploration of this direction. We are actively maintaining a to include the most up-to-date sim-to-real research outcomes to help the researchers in their work.
Authors: Enzhe Lu, Zhejun Jiang, Jingyuan Liu, Yulun Du, Tao Jiang, Chao Hong, Shaowei Liu, Weiran He, Enming Yuan, Yuzhi Wang, Zhiqi Huang, Huan Yuan, Suting Xu, Xinran Xu, Guokun Lai, Yanru Chen, Huabin Zheng, Junjie Yan, Jianlin Su, Yuxin Wu, Neo Y. Zhang, Zhilin Yang, Xinyu Zhou, Mingxing Zhang, Jiezhong Qiu
Abstract: Scaling the effective context length is essential for advancing large language models (LLMs) toward artificial general intelligence (AGI). However, the quadratic increase in computational complexity inherent in traditional attention mechanisms presents a prohibitive overhead. Existing approaches either impose strongly biased structures, such as sink or window attention which are task-specific, or radically modify the attention mechanism into linear approximations, whose performance in complex reasoning tasks remains inadequately explored. In this work, we propose a solution that adheres to the ``less structure'' principle, allowing the model to determine where to attend autonomously, rather than introducing predefined biases. We introduce Mixture of Block Attention (MoBA), an innovative approach that applies the principles of Mixture of Experts (MoE) to the attention mechanism. This novel architecture demonstrates superior performance on long-context tasks while offering a key advantage: the ability to seamlessly transition between full and sparse attention, enhancing efficiency without the risk of compromising performance. MoBA has already been deployed to support Kimi's long-context requests and demonstrates significant advancements in efficient attention computation for LLMs. Our code is available at https://github.com/MoonshotAI/MoBA.
Authors: Junyi Guan, Abhijith Sharma, Chong Tian, Salem Lahlou
Abstract: Spiking Neural Networks (SNNs) are increasingly explored for their energy efficiency and robustness in real-world applications, yet their privacy risks remain largely unexamined. In this work, we investigate the susceptibility of SNNs to Membership Inference Attacks (MIAs) -- a major privacy threat where an adversary attempts to determine whether a given sample was part of the training dataset. While prior work suggests that SNNs may offer inherent robustness due to their discrete, event-driven nature, we find that its resilience diminishes as latency (T) increases. Furthermore, we introduce an input dropout strategy under black box setting, that significantly enhances membership inference in SNNs. Our findings challenge the assumption that SNNs are inherently more secure, and even though they are expected to be better, our results reveal that SNNs exhibit privacy vulnerabilities that are equally comparable to Artificial Neural Networks (ANNs). Our code is available at https://anonymous.4open.science/r/MIA_SNN-3610.
Authors: Dimitrios Troullinos, Georgios Chalkiadakis, Ioannis Papamichail, Markos Papageorgiou
Abstract: In this paper we present a novel approach for multiagent decision making in dynamic environments based on Factor Graphs and the Max-Sum algorithm, considering asynchronous variable reassignments and distributed message-passing among agents. Motivated by the challenging domain of lane-free traffic where automated vehicles can communicate and coordinate as agents, we propose a more realistic communication framework for Factor Graph formulations that satisfies the above-mentioned restrictions, along with Conditional Max-Sum: an extension of Max-Sum with a revised message-passing process that is better suited for asynchronous settings. The overall application in lane-free traffic can be viewed as a hybrid system where the Factor Graph formulation undertakes the strategic decision making of vehicles, that of desired lateral alignment in a coordinated manner; and acts on top of a rule-based method we devise that provides a structured representation of the lane-free environment for the factors, while also handling the underlying control of vehicles regarding core operations and safety. Our experimental evaluation showcases the capabilities of the proposed framework in problems with intense coordination needs when compared to a domain-specific baseline without communication, and an increased adeptness of Conditional Max-Sum with respect to the standard algorithm.
Authors: Manal Rahal, Bestoun S. Ahmed, Gergely Szabados, Torgny Fornstedt, Jorgen Samuelsson
Abstract: Poor data quality limits the advantageous power of Machine Learning (ML) and weakens high-performing ML software systems. Nowadays, data are more prone to the risk of poor quality due to their increasing volume and complexity. Therefore, tedious and time-consuming work goes into data preparation and improvement before moving further in the ML pipeline. To address this challenge, we propose an intelligent data-centric evaluation framework that can identify high-quality data and improve the performance of an ML system. The proposed framework combines the curation of quality measurements and unsupervised learning to distinguish high- and low-quality data. The framework is designed to integrate flexible and general-purpose methods so that it is deployed in various domains and applications. To validate the outcomes of the designed framework, we implemented it in a real-world use case from the field of analytical chemistry, where it is tested on three datasets of anti-sense oligonucleotides. A domain expert is consulted to identify the relevant quality measurements and evaluate the outcomes of the framework. The results show that the quality-centric data evaluation framework identifies the characteristics of high-quality data that guide the conduct of efficient laboratory experiments and consequently improve the performance of the ML system.
Authors: Suresh Babu Nettur, Shanthi Karpurapu, Unnati Nettur, Likhit Sagar Gajja, Sravanthy Myneni, Akhil Dusi
Abstract: GitHub Copilot is transforming software development by automating tasks and boosting productivity through AI-driven code generation. In this paper, we con-duct a literature survey to synthesize insights on Copilot's impact on productivity and security. We review academic journal databases, industry reports, and official docu-mentation to highlight key findings and challenges. While Copilot accelerates coding and prototyping, concerns over security vulnerabilities and intellectual property risks persist. Drawing from the literature, we provide a perspective on best practices and future directions for responsible AI adoption in software engineering, offering action-able insights for developers and organizations to integrate Copilot effectively while maintaining high standards of quality and security.
Authors: Alana Santana, Paula P. Costa, Esther L. Colombini
Abstract: Autonomous artificial agents must be able to learn behaviors in complex environments without humans to design tasks and rewards. Designing these functions for each environment is not feasible, thus, motivating the development of intrinsic reward functions. In this paper, we propose using several cognitive elements that have been neglected for a long time to build an internal world model for an intrinsically motivated agent. Our agent performs satisfactory iterations with the environment, learning complex behaviors without needing previously designed reward functions. We used 18 Atari games to evaluate what cognitive skills emerge in games that require reactive and deliberative behaviors. Our results show superior performance compared to the state-of-the-art in many test cases with dense and sparse rewards.
Authors: Giorgio Franceschelli, Mirco Musolesi
Abstract: Despite the increasing use of large language models for creative tasks, their outputs often lack diversity. Common solutions, such as sampling at higher temperatures, can compromise the quality of the results. Drawing on information theory, we propose a context-based score to quantitatively evaluate value and originality. This score incentivizes accuracy and adherence to the request while fostering divergence from the learned distribution. We propose using our score as a reward in a reinforcement learning framework to fine-tune large language models for maximum performance. We validate our strategy through experiments in poetry generation and math problem solving, demonstrating that it enhances the value and originality of the generated solutions.
Authors: Lee Cohen, Jack Hsieh, Connie Hong, Judy Hanwen Shen
Abstract: In an era of increasingly capable foundation models, job seekers are turning to generative AI tools to enhance their application materials. However, unequal access to and knowledge about generative AI tools can harm both employers and candidates by reducing the accuracy of hiring decisions and giving some candidates an unfair advantage. To address these challenges, we introduce a new variant of the strategic classification framework tailored to manipulations performed using large language models, accommodating varying levels of manipulations and stochastic outcomes. We propose a ``two-ticket'' scheme, where the hiring algorithm applies an additional manipulation to each submitted resume and considers this manipulated version together with the original submitted resume. We establish theoretical guarantees for this scheme, showing improvements for both the fairness and accuracy of hiring decisions when the true positive rate is maximized subject to a no false positives constraint. We further generalize this approach to an $n$-ticket scheme and prove that hiring outcomes converge to a fixed, group-independent decision, eliminating disparities arising from differential LLM access. Finally, we empirically validate our framework and the performance of our two-ticket scheme on real resumes using an open-source resume screening tool.
Authors: Jake C. Snell, Thomas L. Griffiths
Abstract: As machine learning-based prediction systems are increasingly used in high-stakes situations, it is important to understand how such predictive models will perform upon deployment. Distribution-free uncertainty quantification techniques such as conformal prediction provide guarantees about the loss black-box models will incur even when the details of the models are hidden. However, such methods are based on frequentist probability, which unduly limits their applicability. We revisit the central aspects of conformal prediction from a Bayesian perspective and thereby illuminate the shortcomings of frequentist guarantees. We propose a practical alternative based on Bayesian quadrature that provides interpretable guarantees and offers a richer representation of the likely range of losses to be observed at test time.
Authors: Yucheng Shi, Tianze Yang, Canyu Chen, Quanzheng Li, Tianming Liu, Xiang Li, Ninghao Liu
Abstract: Large Language Models (LLMs) have shown remarkable capabilities in general domains but often struggle with tasks requiring specialized knowledge. Conventional Retrieval-Augmented Generation (RAG) techniques typically retrieve external information from static knowledge bases, which can be outdated or incomplete, missing fine-grained clinical details essential for accurate medical question answering. In this work, we propose SearchRAG, a novel framework that overcomes these limitations by leveraging real-time search engines. Our method employs synthetic query generation to convert complex medical questions into search-engine-friendly queries and utilizes uncertainty-based knowledge selection to filter and incorporate the most relevant and informative medical knowledge into the LLM's input. Experimental results demonstrate that our method significantly improves response accuracy in medical question answering tasks, particularly for complex questions requiring detailed and up-to-date knowledge.
Authors: Yen-Siang Wu, Chi-Pin Huang, Fu-En Yang, Yu-Chiang Frank Wang
Abstract: Text-to-video (T2V) diffusion models have shown promising capabilities in synthesizing realistic videos from input text prompts. However, the input text description alone provides limited control over the precise objects movements and camera framing. In this work, we tackle the motion customization problem, where a reference video is provided as motion guidance. While most existing methods choose to fine-tune pre-trained diffusion models to reconstruct the frame differences of the reference video, we observe that such strategy suffer from content leakage from the reference video, and they cannot capture complex motion accurately. To address this issue, we propose MotionMatcher, a motion customization framework that fine-tunes the pre-trained T2V diffusion model at the feature level. Instead of using pixel-level objectives, MotionMatcher compares high-level, spatio-temporal motion features to fine-tune diffusion models, ensuring precise motion learning. For the sake of memory efficiency and accessibility, we utilize a pre-trained T2V diffusion model, which contains considerable prior knowledge about video motion, to compute these motion features. In our experiments, we demonstrate state-of-the-art motion customization performances, validating the design of our framework.
Authors: Hankang Gu, Shangbo Wang, Dongyao Jia, Yuli Zhang, Yanrong Luo, Guoqiang Mao, Jianping Wang, Eng Gee Lim
Abstract: Adaptive Traffic Signal Control (ATSC) has become a popular research topic in intelligent transportation systems. Regional Traffic Signal Control (RTSC) using the Multi-agent Deep Reinforcement Learning (MADRL) technique has become a promising approach for ATSC due to its ability to achieve the optimum trade-off between scalability and optimality. Most existing RTSC approaches partition a traffic network into several disjoint regions, followed by applying centralized reinforcement learning techniques to each region. However, the pursuit of cooperation among RTSC agents still remains an open issue and no communication strategy for RTSC agents has been investigated. In this paper, we propose communication strategies to capture the correlation of micro-traffic states among lanes and the correlation of macro-traffic states among intersections. We first justify the evolution equation of the RTSC process is Markovian via a system of store-and-forward queues. Next, based on the evolution equation, we propose two GAT-Aggregated (GA2) communication modules--GA2-Naive and GA2-Aug to extract both intra-region and inter-region correlations between macro and micro traffic states. While GA2-Naive only considers the movements at each intersection, GA2-Aug also considers the lane-changing behavior of vehicles. Two proposed communication modules are then aggregated into two existing novel RTSC frameworks--RegionLight and Regional-DRL. Experimental results demonstrate that both GA2-Naive and GA2-Aug effectively improve the performance of existing RTSC frameworks under both real and synthetic scenarios. Hyperparameter testing also reveals the robustness and potential of our communication modules in large-scale traffic networks.
Authors: Difan Deng, Marius Lindauer
Abstract: We present Neural Attention Search (NAtS), a framework that automatically evaluates the importance of each token within a sequence and determines if the corresponding token can be dropped after several steps. This approach can efficiently reduce the KV cache sizes required by transformer-based models during inference and thus reduce inference costs. In this paper, we design a search space that contains three token types: (i) Global Tokens will be preserved and queried by all the following tokens. (ii) Local Tokens survive until the next global token appears. (iii) Sliding Window Tokens have an impact on the inference of a fixed size of the next following tokens. Similar to the One-Shot Neural Architecture Search approach, this token-type information can be learned jointly with the architecture weights via a learnable attention mask. Experiments on both training a new transformer from scratch and fine-tuning existing large language models show that NAtS can efficiently reduce the KV cache size required for the models while maintaining the models' performance.
Authors: Danial Abshari, Meera Sridhar
Abstract: In our increasingly interconnected world, Cyber-Physical Systems (CPS) play a crucial role in industries like healthcare, transportation, and manufacturing by combining physical processes with computing power. These systems, however, face many challenges, especially regarding security and system faults. Anomalies in CPS may indicate unexpected problems, from sensor malfunctions to cyber-attacks, and must be detected to prevent failures that can cause harm or disrupt services. This paper provides an overview of the different ways researchers have approached anomaly detection in CPS. We categorize and compare methods like machine learning, deep learning, mathematical models, invariant, and hybrid techniques. Our goal is to help readers understand the strengths and weaknesses of these methods and how they can be used to create safer, more reliable CPS. By identifying the gaps in current solutions, we aim to encourage future research that will make CPS more secure and adaptive in our increasingly automated world.
Authors: Myra Cheng, Sunny Yu, Dan Jurafsky
Abstract: Should LLMs generate language that makes them seem human? Human-like language might improve user experience, but might also lead to overreliance and stereotyping. Assessing these potential impacts requires a systematic way to measure human-like tone in LLM outputs. We introduce HumT and SocioT, metrics for human-like tone and other dimensions of social perceptions in text data based on relative probabilities from an LLM. By measuring HumT across preference and usage datasets, we find that users prefer less human-like outputs from LLMs. HumT also offers insights into the impacts of anthropomorphism: human-like LLM outputs are highly correlated with warmth, social closeness, femininity, and low status, which are closely linked to the aforementioned harms. We introduce DumT, a method using HumT to systematically control and reduce the degree of human-like tone while preserving model performance. DumT offers a practical approach for mitigating risks associated with anthropomorphic language generation.
Authors: Yingqian Cui, Pengfei He, Jingying Zeng, Hui Liu, Xianfeng Tang, Zhenwei Dai, Yan Han, Chen Luo, Jing Huang, Zhen Li, Suhang Wang, Yue Xing, Jiliang Tang, Qi He
Abstract: Chain-of-Thought (CoT) reasoning, which breaks down complex tasks into intermediate reasoning steps, has significantly enhanced the performance of large language models (LLMs) on challenging tasks. However, the detailed reasoning process in CoT often incurs long generation times and high computational costs, partly due to the inclusion of unnecessary steps. To address this, we propose a method to identify critical reasoning steps using perplexity as a measure of their importance: a step is deemed critical if its removal causes a significant increase in perplexity. Our method enables models to focus solely on generating these critical steps. This can be achieved through two approaches: refining demonstration examples in few-shot CoT or fine-tuning the model using selected examples that include only critical steps. Comprehensive experiments validate the effectiveness of our method, which achieves a better balance between the reasoning accuracy and efficiency of CoT.
Authors: Khaled Mohammed Saifuddin, Jonathan Shihao Ji, Esra Akbas
Abstract: Recent advancements in Graph Contrastive Learning (GCL) have demonstrated remarkable effectiveness in improving graph representations. However, relying on predefined augmentations (e.g., node dropping, edge perturbation, attribute masking) may result in the loss of task-relevant information and a lack of adaptability to diverse input data. Furthermore, the selection of negative samples remains rarely explored. In this paper, we introduce HyperGCL, a novel multimodal GCL framework from a hypergraph perspective. HyperGCL constructs three distinct hypergraph views by jointly utilizing the input graph's structure and attributes, enabling a comprehensive integration of multiple modalities in contrastive learning. A learnable adaptive topology augmentation technique enhances these views by preserving important relations and filtering out noise. View-specific encoders capture essential characteristics from each view, while a network-aware contrastive loss leverages the underlying topology to define positive and negative samples effectively. Extensive experiments on benchmark datasets demonstrate that HyperGCL achieves state-of-the-art node classification performance.
Authors: Sirisha Velampalli, Chandrashekar Muniyappa, Ashutosh Saxena
Abstract: Emojis are being frequently used in todays digital world to express from simple to complex thoughts more than ever before. Hence, they are also being used in sentiment analysis and targeted marketing campaigns. In this work, we performed sentiment analysis of Tweets as well as on emoji dataset from the Kaggle. Since tweets are sentences we have used Universal Sentence Encoder (USE) and Sentence Bidirectional Encoder Representations from Transformers (SBERT) end-to-end sentence embedding models to generate the embeddings which are used to train the Standard fully connected Neural Networks (NN), and LSTM NN models. We observe the text classification accuracy was almost the same for both the models around 98 percent. On the contrary, when the validation set was built using emojis that were not present in the training set then the accuracy of both the models reduced drastically to 70 percent. In addition, the models were also trained using the distributed training approach instead of a traditional singlethreaded model for better scalability. Using the distributed training approach, we were able to reduce the run-time by roughly 15% without compromising on accuracy. Finally, as part of explainable AI the Shap algorithm was used to explain the model behaviour and check for model biases for the given feature set.
Authors: Sachini Weerasekara, Sagar Kamarthi, Jacqueline Isaacs
Abstract: Predicting medical events in advance within critical care settings is paramount for patient outcomes and resource management. Utilizing predictive models, healthcare providers can anticipate issues such as cardiac arrest, sepsis, or respiratory failure before they manifest. Recently, there has been a surge in research focusing on forecasting adverse medical event onsets prior to clinical manifestation using machine learning. However, while these models provide temporal prognostic predictions for the occurrence of a specific adverse event of interest within defined time intervals, their interpretability often remains a challenge. In this work, we explore the applicability of neural temporal point processes in the context of adverse event onset prediction, with the aim of explaining clinical pathways and providing interpretable insights. Our experiments span six state-of-the-art neural point processes and six critical care datasets, each focusing on the onset of distinct adverse events. This work represents a novel application class of neural temporal point processes in event prediction.
Authors: Yunpeng Xiao, Youpeng Zhao, Kai Shu
Abstract: Natural language understanding (NLU) is a task that enables machines to understand human language. Some tasks, such as stance detection and sentiment analysis, are closely related to individual subjective perspectives, thus termed individual-level NLU. Previously, these tasks are often simplified to text-level NLU tasks, ignoring individual factors. This not only makes inference difficult and unexplainable but often results in a large number of label errors when creating datasets. To address the above limitations, we propose a new NLU annotation guideline based on individual-level factors. Specifically, we incorporate other posts by the same individual and then annotate individual subjective perspectives after considering all individual posts. We use this guideline to expand and re-annotate the stance detection and topic-based sentiment analysis datasets. We find that error rates in the samples were as high as 31.7\% and 23.3\%. We further use large language models to conduct experiments on the re-annotation datasets and find that the large language models perform well on both datasets after adding individual factors. Both GPT-4o and Llama3-70B can achieve an accuracy greater than 87\% on the re-annotation datasets. We also verify the effectiveness of individual factors through ablation studies. We call on future researchers to add individual factors when creating such datasets. Our re-annotation dataset can be found at https://github.com/24yearsoldstudent/Individual-NLU
Authors: Jian Wang, Yinpei Dai, Yichi Zhang, Ziqiao Ma, Wenjie Li, Joyce Chai
Abstract: Intelligent tutoring agents powered by large language models (LLMs) have been increasingly explored to deliver personalized guidance in areas such as language learning and science education. However, their capabilities in guiding users to solve complex real-world tasks remain underexplored. To address this limitation, in this work, we focus on coding tutoring, a challenging problem that requires tutors to proactively guide students toward completing predefined coding tasks. We propose a novel agent workflow, Trace-and-Verify (TRAVER), which combines knowledge tracing to estimate a student's knowledge state and turn-by-turn verification to ensure effective guidance toward task completion. We introduce DICT, an automatic evaluation protocol that assesses tutor agents holistically using controlled student simulation and code generation tests. Extensive experiments reveal the challenges of coding tutoring and demonstrate that TRAVER achieves a significantly higher success rate. Although we use code tutoring as an example in this paper, our results and findings can be extended beyond coding, providing valuable insights into advancing tutoring agents for a variety of tasks.
Authors: Tejas Srinivasan, Jesse Thomason
Abstract: Trust biases how users rely on AI recommendations in AI-assisted decision-making tasks, with low and high levels of trust resulting in increased under- and over-reliance, respectively. We propose that AI assistants should adapt their behavior through trust-adaptive interventions to mitigate such inappropriate reliance. For instance, when user trust is low, providing an explanation can elicit more careful consideration of the assistant's advice by the user. In two decision-making scenarios -- laypeople answering science questions and doctors making medical diagnoses -- we find that providing supporting and counter-explanations during moments of low and high trust, respectively, yields up to 38% reduction in inappropriate reliance and 20% improvement in decision accuracy. We are similarly able to reduce over-reliance by adaptively inserting forced pauses to promote deliberation. Our results highlight how AI adaptation to user trust facilitates appropriate reliance, presenting exciting avenues for improving human-AI collaboration.
Authors: Dhananjay Ashok, Jonathan May
Abstract: Autoregressive Language Models output text by sequentially predicting the next token to generate, with modern methods like Chain-of-Thought (CoT) prompting achieving state-of-the-art reasoning capabilities by scaling the number of generated tokens. However, are there times when we can infer how the model will behave (e.g. abstain from answering a question) early in the computation, making generation unnecessary? We show that internal representation of input tokens alone can often precisely predict, not just the next token, but eventual behavior over the entire output sequence. We leverage this capacity and learn probes on internal states to create early warning (and exit) systems. Specifically, if the probes can confidently estimate the way the LM is going to behave, then the system will avoid generating tokens altogether and return the estimated behavior instead. On 27 text classification datasets spanning five different tasks, we apply this method to estimate the eventual answer of an LM under CoT prompting, reducing inference costs by 65% (average) while suffering an accuracy loss of no more than 1.4% (worst case). We demonstrate the potential of this method to pre-emptively identify when a model will abstain from answering a question, fail to follow output format specifications, or give a low-confidence response. We explore the limits of this capability, showing that probes generalize to unseen datasets, but perform worse when LM outputs are longer and struggle to predict properties that require access to knowledge that the models themselves lack. Encouragingly, performance scales with model size, suggesting applicability to the largest of models
Authors: Chenyan Zhao, Mariana Silva, Seth Poulsen
Abstract: Providing evaluations to student work is a critical component of effective student learning, and automating its process can significantly reduce the workload on human graders. Automatic Short Answer Grading (ASAG) systems, enabled by advancements in Large Language Models (LLMs), offer a promising solution for assessing and providing instant feedback for open-ended student responses. In this paper, we present an ASAG pipeline leveraging state-of-the-art LLMs. Our new LLM-based ASAG pipeline achieves better performances than existing custom-built models on the same datasets. We also compare the grading performance of three OpenAI models: GPT-4, GPT-4o, and o1-preview. Our results demonstrate that GPT-4o achieves the best balance between accuracy and cost-effectiveness. On the other hand, o1-preview, despite higher accuracy, exhibits a larger variance in error that makes it less practical for classroom use. We investigate the effects of incorporating instructor-graded examples into prompts using no examples, random selection, and Retrieval-Augmented Generation (RAG)-based selection strategies. Our findings indicate that providing graded examples enhances grading accuracy, with RAG-based selection outperforming random selection. Additionally, integrating grading rubrics improves accuracy by offering a structured standard for evaluation.
Authors: Xingyue Huang, Pablo Barcel\'o, Michael M. Bronstein, \.Ismail \.Ilkan Ceylan, Mikhail Galkin, Juan L Reutter, Miguel Romero Orth
Abstract: Knowledge Graph Foundation Models (KGFMs) are at the frontier for deep learning on knowledge graphs (KGs), as they can generalize to completely novel knowledge graphs with different relational vocabularies. Despite their empirical success, our theoretical understanding of KGFMs remains very limited. In this paper, we conduct a rigorous study of the expressive power of KGFMs. Specifically, we show that the expressive power of KGFMs directly depends on the motifs that are used to learn the relation representations. We then observe that the most typical motifs used in the existing literature are binary, as the representations are learned based on how pairs of relations interact, which limits the model's expressiveness. As part of our study, we design more expressive KGFMs using richer motifs, which necessitate learning relation representations based on, e.g., how triples of relations interact with each other. Finally, we empirically validate our theoretical findings, showing that the use of richer motifs results in better performance on a wide range of datasets drawn from different domains.
Authors: Liangqi Lei, Keke Gai, Jing Yu, Liehuang Zhu, Qi Wu
Abstract: Latent diffusion models have exhibited considerable potential in generative tasks. Watermarking is considered to be an alternative to safeguard the copyright of generative models and prevent their misuse. However, in the context of model distribution scenarios, the accessibility of models to large scale of model users brings new challenges to the security, efficiency and robustness of existing watermark solutions. To address these issues, we propose a secure and efficient watermarking solution. A new security mechanism is designed to prevent watermark leakage and watermark escape, which considers watermark randomness and watermark-model association as two constraints for mandatory watermark injection. To reduce the time cost of training the security module, watermark injection and the security mechanism are decoupled, ensuring that fine-tuning VAE only accomplishes the security mechanism without the burden of learning watermark patterns. A watermark distribution-based verification strategy is proposed to enhance the robustness against diverse attacks in the model distribution scenarios. Experimental results prove that our watermarking consistently outperforms existing six baselines on effectiveness and robustness against ten image processing attacks and adversarial attacks, while enhancing security in the distribution scenarios.
Authors: Sichu Liang, Linhai Zhang, Hongyu Zhu, Wenwen Wang, Yulan He, Deyu Zhou
Abstract: Medical question answering requires extensive access to specialized conceptual knowledge. The current paradigm, Retrieval-Augmented Generation (RAG), acquires expertise medical knowledge through large-scale corpus retrieval and uses this knowledge to guide a general-purpose large language model (LLM) for generating answers. However, existing retrieval approaches often overlook the importance of factual knowledge, which limits the relevance of retrieved conceptual knowledge and restricts its applicability in real-world scenarios, such as clinical decision-making based on Electronic Health Records (EHRs). This paper introduces RGAR, a recurrence generation-augmented retrieval framework that retrieves both relevant factual and conceptual knowledge from dual sources (i.e., EHRs and the corpus), allowing them to interact and refine each another. Through extensive evaluation across three factual-aware medical question answering benchmarks, RGAR establishes a new state-of-the-art performance among medical RAG systems. Notably, the Llama-3.1-8B-Instruct model with RGAR surpasses the considerably larger, RAG-enhanced GPT-3.5. Our findings demonstrate the benefit of extracting factual knowledge for retrieval, which consistently yields improved generation quality.
Authors: Ameesh Shah, Niklas Lauffer, Thomas Chen, Nikhil Pitta, Sanjit A. Seshia
Abstract: One approach for improving sample efficiency in cooperative multi-agent learning is to decompose overall tasks into sub-tasks that can be assigned to individual agents. We study this problem in the context of reward machines: symbolic tasks that can be formally decomposed into sub-tasks. In order to handle settings without a priori knowledge of the environment, we introduce a framework that can learn the optimal decomposition from model-free interactions with the environment. Our method uses a task-conditioned architecture to simultaneously learn an optimal decomposition and the corresponding agents' policies for each sub-task. In doing so, we remove the need for a human to manually design the optimal decomposition while maintaining the sample-efficiency benefits of improved credit assignment. We provide experimental results in several deep reinforcement learning settings, demonstrating the efficacy of our approach. Our results indicate that our approach succeeds even in environments with codependent agent dynamics, enabling synchronous multi-agent learning not achievable in previous works.
Authors: Vishal Dey, Xiao Hu, Xia Ning
Abstract: Despite recent advancements, most computational methods for molecule optimization are constrained to single- or double-property optimization tasks and suffer from poor scalability and generalizability to novel optimization tasks. Meanwhile, Large Language Models (LLMs) demonstrate remarkable out-of-domain generalizability to novel tasks. To demonstrate LLMs' potential for molecule optimization, we introduce $\mathtt{MoMUInstruct}$, the first high-quality instruction-tuning dataset specifically focused on complex multi-property molecule optimization tasks. Leveraging $\mathtt{MoMUInstruct}$, we develop $\mathtt{GeLLM^3O}$s, a series of instruction-tuned LLMs for molecule optimization. Extensive evaluations across 5 in-domain and 5 out-of-domain tasks demonstrate that $\mathtt{GeLLM^3O}$s consistently outperform state-of-the-art baselines. $\mathtt{GeLLM^3O}$s also exhibit outstanding zero-shot generalization to unseen tasks, significantly outperforming powerful closed-source LLMs. Such strong generalizability demonstrates the tremendous potential of $\mathtt{GeLLM^3O}$s as foundational models for molecule optimization, thereby tackling novel optimization tasks without resource-intensive retraining. $\mathtt{MoMUInstruct}$, models, and code are accessible through https://github.com/ninglab/GeLLMO.
Authors: Vince Kurtz, Joel W. Burdick
Abstract: Generative control policies have recently unlocked major progress in robotics. These methods produce action sequences via diffusion or flow matching, with training data provided by demonstrations. But despite enjoying considerable success on difficult manipulation problems, generative policies come with two key limitations. First, behavior cloning requires expert demonstrations, which can be time-consuming and expensive to obtain. Second, existing methods are limited to relatively slow, quasi-static tasks. In this paper, we leverage a tight connection between sampling-based predictive control and generative modeling to address each of these issues. In particular, we introduce generative predictive control, a supervised learning framework for tasks with fast dynamics that are easy to simulate but difficult to demonstrate. We then show how trained flow-matching policies can be warm-started at run-time, maintaining temporal consistency and enabling fast feedback rates. We believe that generative predictive control offers a complementary approach to existing behavior cloning methods, and hope that it paves the way toward generalist policies that extend beyond quasi-static demonstration-oriented tasks.
Authors: Ziyuan Liu, Ruifei Zhu, Long Gao, Yuanxiu Zhou, Jingyu Ma, Yuantao Gu
Abstract: Deep learning has achieved significant success in the field of remote sensing image change detection (CD), yet two major challenges remain: the scarcity of sub-meter, all-inclusive open-source CD datasets, and the difficulty of achieving consistent and satisfactory detection results across images with varying change areas. To address these issues, we introduce the JL1-CD dataset, which contains 5,000 pairs of 512 x 512 pixel images with a resolution of 0.5 to 0.75 meters. Additionally, we propose a multi-teacher knowledge distillation (MTKD) framework for CD. Experimental results on the JL1-CD and SYSU-CD datasets demonstrate that the MTKD framework significantly improves the performance of CD models with various network architectures and parameter sizes, achieving new state-of-the-art results. The code is available at https://github.com/circleLZY/MTKD-CD.
Authors: Siddarth Srinivasan, Ezra Karger, Michiel Bakker, Yiling Chen
Abstract: Common sense suggests that when individuals explain why they believe something, we can arrive at more accurate conclusions than when they simply state what they believe. Yet, there is no known mechanism that provides incentives to elicit explanations for beliefs from agents. This likely stems from the fact that standard Bayesian models make assumptions (like conditional independence of signals) that preempt the need for explanations, in order to show efficient information aggregation. A natural justification for the value of explanations is that agents' beliefs tend to be drawn from overlapping sources of information, so agents' belief reports do not reveal all that needs to be known. Indeed, this work argues that rationales-explanations of an agent's private information-lead to more efficient aggregation by allowing agents to efficiently identify what information they share and what information is new. Building on this model of rationales, we present a novel 'deliberation mechanism' to elicit rationales from agents in which truthful reporting of beliefs and rationales is a perfect Bayesian equilibrium.
Authors: Yanbang Sun, Qing Huang, Xiaoxue Ren, Zhenchang Xing, Xiaohong Li, Junjie Wang
Abstract: The API Knowledge Graph (API KG) is a structured network that models API entities and their relations, providing essential semantic insights for tasks such as API recommendation, code generation, and API misuse detection. However, constructing a knowledge-rich and reliable API KG presents several challenges. Existing schema-based methods rely heavily on manual annotations to design KG schemas, leading to excessive manual overhead. On the other hand, schema-free methods, due to the lack of schema guidance, are prone to introducing noise, reducing the KG's reliability. To address these issues, we propose the Explore-Construct-Filter framework, an automated approach for API KG construction based on large language models (LLMs). This framework consists of three key modules: 1) KG exploration: LLMs simulate the workflow of annotators to automatically design a schema with comprehensive type triples, minimizing human intervention; 2) KG construction: Guided by the schema, LLMs extract instance triples to construct a rich yet unreliable API KG; 3) KG filtering: Removing invalid type triples and suspicious instance triples to construct a rich and reliable API KG. Experimental results demonstrate that our method surpasses the state-of-the-art method, achieving a 25.2% improvement in F1 score. Moreover, the Explore-Construct-Filter framework proves effective, with the KG exploration module increasing KG richness by 133.6% and the KG filtering module improving reliability by 26.6%. Finally, cross-model experiments confirm the generalizability of our framework.
Authors: Yifei Xu, Tusher Chakraborty, Emre K{\i}c{\i}man, Bibek Aryal, Eduardo Rodrigues, Srinagesh Sharma, Roberto Estevao, Maria Angels de Luis Balaguer, Jessica Wolk, Rafael Padilha, Leonardo Nunes, Shobana Balakrishnan, Songwu Lu, Ranveer Chandra
Abstract: Fine-tuning large language models (LLMs) to align with user preferences is challenging due to the high cost of quality human annotations in Reinforcement Learning from Human Feedback (RLHF) and the generalizability limitations of AI Feedback. To address these challenges, we propose RLTHF, a human-AI hybrid framework that combines LLM-based initial alignment with selective human annotations to achieve full-human annotation alignment with minimal effort. RLTHF identifies hard-to-annotate samples mislabeled by LLMs using a reward model's reward distribution and iteratively enhances alignment by integrating strategic human corrections while leveraging LLM's correctly labeled samples. Evaluations on HH-RLHF and TL;DR datasets show that RLTHF reaches full-human annotation-level alignment with only 6-7% of the human annotation effort. Furthermore, models trained on RLTHF's curated datasets for downstream tasks outperform those trained on fully human-annotated datasets, underscoring the effectiveness of RLTHF's strategic data curation.
Authors: Yuxiang Wang, Junhao Gan, Jianzhong Qi
Abstract: Question answering on free-form tables (TableQA) is challenging due to the absence of predefined schemas and the presence of noise in large tables. While Large Language Models (LLMs) have shown promise in TableQA, they struggle with large free-form tables and noise sensitivity. To address these challenges, we propose TabSD, a SQL-based decomposition model that enhances LLMs' ability to process large free-form tables. TabSD generates SQL queries to guide the table decomposition, remove noise, and processes sub-tables for better answer generation. Additionally, SQL Verifier refines SQL outputs to enhance decomposition accuracy. We introduce two TableQA datasets with large free-form tables, SLQA and SEQA, which consist solely of large free-form tables and will be publicly available. Experimental results on four benchmark datasets demonstrate that TABSD outperforms the best-existing baseline models by 23.07%, 2.84%, 23.24% and 9.32% in accuracy, respectively, highlighting its effectiveness in handling large and noisy free-form tables.
Authors: Guanming Xiong, Haochen Li, Wen Zhao
Abstract: This study explores how to enhance the reasoning capabilities of large language models (LLMs) in knowledge base question answering (KBQA) by leveraging Monte Carlo Tree Search (MCTS). Semantic parsing-based KBQA methods are particularly challenging as these approaches require locating elements from knowledge bases and generating logical forms, demanding not only extensive annotated data but also strong reasoning capabilities. Although recent approaches leveraging LLMs as agents have demonstrated considerable potential, these studies are inherently constrained by their linear decision-making processes. To address this limitation, we propose a MCTS-based framework that enhances LLMs' reasoning capabilities through tree search methodology. We design a carefully designed step-wise reward mechanism that requires only direct prompting of open-source instruction LLMs without additional fine-tuning. Experimental results demonstrate that our approach significantly outperforms linear decision-making methods, particularly in low-resource scenarios. Additionally, we contribute new data resources to the KBQA community by annotating intermediate reasoning processes for existing question-SPARQL datasets using distant supervision. Experimental results on the extended dataset demonstrate that our method achieves comparable performance to fully supervised models while using significantly less training data.
Authors: Simen Hexeberg, Mandar Chitre, Matthias Hoffmann-Kuhnt, Bing Wen Low
Abstract: Changes in bird populations can indicate broader changes in ecosystems, making birds one of the most important animal groups to monitor. Combining machine learning and passive acoustics enables continuous monitoring over extended periods without direct human involvement. However, most existing techniques require extensive expert-labeled datasets for training and cannot easily detect time-overlapping calls in busy soundscapes. We propose a semi-supervised acoustic bird detector designed to allow both the detection of time-overlapping calls (when separated in frequency) and the use of few labeled training samples. The classifier is trained and evaluated on a combination of community-recorded open-source data and long-duration soundscape recordings from Singapore. It achieves a mean F0.5 score of 0.701 across 315 classes from 110 bird species on a hold-out test set, with an average of 11 labeled training samples per class. It outperforms the state-of-the-art BirdNET classifier on a test set of 103 bird species despite significantly fewer labeled training samples. The detector is further tested on 144 microphone-hours of continuous soundscape data. The rich soundscape in Singapore makes suppression of false positives a challenge on raw, continuous data streams. Nevertheless, we demonstrate that achieving high precision in such environments with minimal labeled training data is possible.
Authors: Yutao Sun, Mingshuai Chen, Tiancheng Zhao, Ruochen Xu, Zilun Zhang, Jianwei Yin
Abstract: Self-improving large language models (LLMs) -- i.e., to improve the performance of an LLM by fine-tuning it with synthetic data generated by itself -- is a promising way to advance the capabilities of LLMs while avoiding extensive supervision. Existing approaches to self-improvement often rely on external supervision signals in the form of seed data and/or assistance from third-party models. This paper presents Crescent -- a simple yet effective framework for generating high-quality synthetic question-answer data in a fully autonomous manner. Crescent first elicits the LLM to generate raw questions via a bait prompt, then diversifies these questions leveraging a rejection sampling-based self-deduplication, and finally feeds the questions to the LLM and collects the corresponding answers by means of majority voting. We show that Crescent sheds light on the potential of true self-improvement with zero external supervision signals for math reasoning; in particular, Crescent-generated question-answer pairs suffice to (i) improve the reasoning capabilities of an LLM while preserving its general performance (especially in the 0-shot setting); and (ii) distil LLM knowledge to weaker models more effectively than existing methods based on seed-dataset augmentation.
Authors: Jialin Ouyang
Abstract: Large language models (LLMs) now achieve near-human performance on standard math word problem benchmarks (e.g., GSM8K), yet their true reasoning ability remains disputed. A key concern is that models often produce confident, yet unfounded, answers to unanswerable problems. We introduce TreeCut, a synthetic dataset that systematically generates infinite unanswerable math word problems and their answerable counterparts, by representing each question as a tree and removing chosen necessary conditions. Experiments show TreeCut effectively induce hallucinations in large language models, including GPT-4o and o3-mini, with rates of 61% and 42% in their respective worst-case scenarios. Further analysis highlights that deeper or more complex trees, composite item names, and removing necessary condition near the middle of a path all increase the likelihood of hallucinations, underscoring the persistent challenges LLMs face in identifying unanswerable math problems.
Authors: Gautham Govind Anil, Sachin Yadav, Dheeraj Nagaraj, Karthikeyan Shanmugam, Prateek Jain
Abstract: We introduce Interleaved Gibbs Diffusion (IGD), a novel generative modeling framework for mixed continuous-discrete data, focusing on constrained generation problems. Prior works on discrete and continuous-discrete diffusion models assume factorized denoising distribution for fast generation, which can hinder the modeling of strong dependencies between random variables encountered in constrained generation. IGD moves beyond this by interleaving continuous and discrete denoising algorithms via a discrete time Gibbs sampling type Markov chain. IGD provides flexibility in the choice of denoisers, allows conditional generation via state-space doubling and inference time scaling via the ReDeNoise method. Empirical evaluations on three challenging tasks-solving 3-SAT, generating molecule structures, and generating layouts-demonstrate state-of-the-art performance. Notably, IGD achieves a 7% improvement on 3-SAT out of the box and achieves state-of-the-art results in molecule generation without relying on equivariant diffusion or domain-specific architectures. We explore a wide range of modeling, and interleaving strategies along with hyperparameters in each of these problems.
Authors: Xiaofei Wen, Wenxuan Zhou, Wenjie Jacky Mo, Muhao Chen
Abstract: Ensuring the safety of large language models (LLMs) is critical as they are deployed in real-world applications. Existing guardrails rely on rule-based filtering or single-pass classification, limiting their ability to handle nuanced safety violations. To address this, we propose ThinkGuard, a critique-augmented guardrail model that distills knowledge from high-capacity LLMs by generating structured critiques alongside safety labels. Fine-tuned on critique-augmented data, the captured deliberative thinking ability drastically enhances the guardrail's cautiousness and interpretability. Evaluated on multiple safety benchmarks, ThinkGuard achieves the highest average F1 and AUPRC, outperforming all baselines. Compared to LLaMA Guard 3, ThinkGuard improves accuracy by 16.1% and macro F1 by 27.0%. Moreover, it surpasses label-only fine-tuned models, confirming that structured critiques enhance both classification precision and nuanced safety reasoning while maintaining computational efficiency.
Authors: Wanqing Cui, Keping Bi, Jiafeng Guo, Xueqi Cheng
Abstract: Commonsense plausibility estimation is critical for evaluating language models (LMs), yet existing generative approaches--reliant on likelihoods or verbalized judgments--struggle with fine-grained discrimination. In this paper, we propose ComPaSS, a novel discriminative framework that quantifies commonsense plausibility by measuring semantic shifts when augmenting sentences with commonsense-related information. Plausible augmentations induce minimal shifts in semantics, while implausible ones result in substantial deviations. Evaluations on two types of fine-grained commonsense plausibility estimation tasks across different backbones, including LLMs and vision-language models (VLMs), show that ComPaSS consistently outperforms baselines. It demonstrates the advantage of discriminative approaches over generative methods in fine-grained commonsense plausibility evaluation. Experiments also show that (1) VLMs yield superior performance to LMs, when integrated with ComPaSS, on vision-grounded commonsense tasks. (2) contrastive pre-training sharpens backbone models' ability to capture semantic nuances, thereby further enhancing ComPaSS.
Authors: Hongjin Qian, Zheng Liu, Chao Gao, Yankai Wang, Defu Lian, Zhicheng Dou
Abstract: In real-world information-seeking scenarios, users have dynamic and diverse needs, requiring RAG systems to demonstrate adaptable resilience. To comprehensively evaluate the resilience of current RAG methods, we introduce HawkBench, a human-labeled, multi-domain benchmark designed to rigorously assess RAG performance across categorized task types. By stratifying tasks based on information-seeking behaviors, HawkBench provides a systematic evaluation of how well RAG systems adapt to diverse user needs. Unlike existing benchmarks, which focus primarily on specific task types (mostly factoid queries) and rely on varying knowledge bases, HawkBench offers: (1) systematic task stratification to cover a broad range of query types, including both factoid and rationale queries, (2) integration of multi-domain corpora across all task types to mitigate corpus bias, and (3) rigorous annotation for high-quality evaluation. HawkBench includes 1,600 high-quality test samples, evenly distributed across domains and task types. Using this benchmark, we evaluate representative RAG methods, analyzing their performance in terms of answer quality and response latency. Our findings highlight the need for dynamic task strategies that integrate decision-making, query interpretation, and global knowledge understanding to improve RAG generalizability. We believe HawkBench serves as a pivotal benchmark for advancing the resilience of RAG methods and their ability to achieve general-purpose information seeking.
Authors: Phaphontee Yamchote, Saw Nay Htet Win, Chainarong Amornbunchornvej, Thanapon Noraset
Abstract: Feature interaction is crucial in predictive machine learning models, as it captures the relationships between features that influence model performance. In this work, we focus on pairwise interactions and investigate their importance in constructing feature graphs for Graph Neural Networks (GNNs). Rather than proposing new methods, we leverage existing GNN models and tools to explore the relationship between feature graph structures and their effectiveness in modeling interactions. Through experiments on synthesized datasets, we uncover that edges between interacting features are important for enabling GNNs to model feature interactions effectively. We also observe that including non-interaction edges can act as noise, degrading model performance. Furthermore, we provide theoretical support for sparse feature graph selection using the Minimum Description Length (MDL) principle. We prove that feature graphs retaining only necessary interaction edges yield a more efficient and interpretable representation than complete graphs, aligning with Occam's Razor. Our findings offer both theoretical insights and practical guidelines for designing feature graphs that improve the performance and interpretability of GNN models.
Authors: Haun Leung, ZiNan Wang
Abstract: It is popular lately to train large language models to be used as chat assistants, but in the conversation between the user and the chat assistant, there are prompts, require multi-turns between the chat assistant and the user. However, there are a number of issues with the multi-turns conversation: The response of the chat assistant is prone to errors and cannot help users achieve their goals; It is difficult for chat assistant to generate responses with different processes based on actual needs for the same command or request; Chat assistant require the use of tools, but the current approach is not elegant and efficient, and the number of tool calls that can be supported is limited. The main reason for these issues is that large language models do not have the thinking ability as a human, lack the reasoning ability and planning ability, and lack the ability to execute plans. To solve these issues, we propose a thinking method based on a built-in chain of thought: In the multi-turns conversation, for each user prompt, the large language model thinks based on elements such as chat history, thinking context, action calls, memory and knowledge, makes detailed reasoning and planning, and actions according to the plan. We also explored how the large language model enhances thinking ability through this thinking method: Collect training datasets according to the thinking method and fine tune the large language model through supervised learning; Train a consistency reward model and use it as a reward function to fine tune the large language model using reinforcement learning, and the reinforced large language model outputs according to this way of thinking. Our experimental results show that the reasoning ability and planning ability of the large language model are enhanced, and the issues in the multi-turns conversation are solved.
Authors: Peiran Wang, Haibing Li, Fu Haohan, Shiyong Li, Yanpeng Wang, Dou Shen
Abstract: In this paper, we introduce an efficient and money-saving automatic parallel strategies search framework on heterogeneous GPUs: Astra. First, Astra searches for the efficiency-optimal parallel strategy in both GPU configurations search space (GPU types and GPU numbers) and parallel parameters search space. Then, Astra also provides the solution on heterogeneous GPUs by mathematically modeling the time consumption of heterogeneous training. At last, Astra is the first to propose the automatic parallel strategy search on money-saving. The experiment results demonstrate that Astra can achieve better throughput than expert-designed strategies. The search time cost for Astra can also be limited to 1.27 seconds in a single-GPU setting and less than 1.35 minutes in a heterogeneous-GPU setting on average with an accuracy of over 95%.
Authors: Chen-An Li, Tzu-Han Lin, Yun-Nung Chen, Hung-yi Lee
Abstract: Large vision-language models (LVLMs) perform outstandingly across various multimodal tasks. However, their ability to evaluate generated content remains limited, and training vision-language reward models (VLRMs) with preference data is computationally expensive. This paper explores a training-free alternative by merging text-based reward models (RMs) with LVLMs to create VLRMs. Our approach shows that integrating these models leads to improved performance over LVLMs' scoring and text-based RMs, offering an efficient method for incorporating textual preferences into LVLMs.
Authors: Peiran Wang, Yang Liu, Yunfei Lu, Jue Hong, Ye Wu
Abstract: Large language model (LLM) systems suffer from the models' unstable ability to generate valid and factual content, resulting in hallucination generation. Current hallucination detection methods heavily rely on out-of-model information sources, such as RAG to assist the detection, thus bringing heavy additional latency. Recently, internal states of LLMs' inference have been widely used in numerous research works, such as prompt injection detection, etc. Considering the interpretability of LLM internal states and the fact that they do not require external information sources, we introduce such states into LLM hallucination detection. In this paper, we systematically analyze different internal states' revealing features during inference forward and comprehensively evaluate their ability in hallucination detection. Specifically, we cut the forward process of a large language model into three stages: understanding, query, generation, and extracting the internal state from these stages. By analyzing these states, we provide a deep understanding of why the hallucinated content is generated and what happened in the internal state of the models. Then, we introduce these internal states into hallucination detection and conduct comprehensive experiments to discuss the advantages and limitations.
Authors: Piyawat Lertvittayakumjorn, David Kinney, Vinodkumar Prabhakaran, Donald Martin, Sunipa Dev
Abstract: Generative large language models (LLMs) have been demonstrated to have gaps in diverse, cultural knowledge across the globe. We investigate the effect of retrieval augmented generation and search-grounding techniques on the ability of LLMs to display familiarity with a diverse range of national cultures. Specifically, we compare the performance of standard LLMs, LLMs augmented with retrievals from a bespoke knowledge base (i.e., KB grounding), and LLMs augmented with retrievals from a web search (i.e., search grounding) on a series of cultural familiarity benchmarks. We find that search grounding significantly improves the LLM performance on multiple-choice benchmarks that test propositional knowledge (e.g., the norms, artifacts, and institutions of national cultures), while KB grounding's effectiveness is limited by inadequate knowledge base coverage and a suboptimal retriever. However, search grounding also increases the risk of stereotypical judgments by language models, while failing to improve evaluators' judgments of cultural familiarity in a human evaluation with adequate statistical power. These results highlight the distinction between propositional knowledge about a culture and open-ended cultural fluency when it comes to evaluating the cultural familiarity of generative LLMs.
Authors: Ziwei Chen, Jiawen Shen, Luna, Kristen Vaccaro
Abstract: Recent work has highlighted the risks of LLM-generated content for a wide range of harmful behaviors, including incorrect and harmful code. In this work, we extend this by studying whether LLM-generated web design contains dark patterns. This work evaluated designs of ecommerce web components generated by four popular LLMs: Claude, GPT, Gemini, and Llama. We tested 13 commonly used ecommerce components (e.g., search, product reviews) and used them as prompts to generate a total of 312 components across all models. Over one-third of generated components contain at least one dark pattern. The majority of dark pattern strategies involve hiding crucial information, limiting users' actions, and manipulating them into making decisions through a sense of urgency. Dark patterns are also more frequently produced in components that are related to company interests. These findings highlight the need for interventions to prevent dark patterns during front-end code generation with LLMs and emphasize the importance of expanding ethical design education to a broader audience.
Authors: Burc Gokden
Abstract: We show that Large Language Model from Power Law Decoder Representations (PLDR-LLM) is a foundational model whose deductive outputs are invariant tensors up to a small perturbation. PLDR-LLM learns a singularity condition for the deductive outputs that enable the once-inferred energy-curvature tensor $\mathbf{G}_{LM}$ to replace the deep neural network of power law graph attention (PLGA) generating the deductive outputs at inference. We demonstrate that a cache for $\mathbf{G}_{LM}$ (G-cache) and KV-cache can be implemented in a straightforward manner to improve the inference time. The invariance and generalizable nature of deductive outputs is at a very high fidelity where deductive outputs have same RMSE and determinant values up to 15 decimal places after caching, and zero-shot benchmark scores remain unchanged. Ablation studies show that learned deductive outputs have distinct loss and accuracy characteristics from models pretrained with transferred, randomly initialized or identity tensors as a constant tensor operator and an LLM with scaled-dot product attention (SDPA) is a special case of PLDR-LLM where $\mathbf{G}_{LM}$ is predefined as identity. The observed invariance characteristic introduces a novel asymmetry between training and inference phases with caching. We outline observed common characteristics of the deductive outputs for the learned singularity condition. We provide an implementation of a training and inference framework for PLDR-LLM with KV-cache and G-cache.
Authors: Shuai Niu, Jing Ma, Hongzhan Lin, Liang Bai, Zhihua Wang, Wei Bi, Yida Xu, Guo Li, Xian Yang
Abstract: Large language models (LLMs) have shown remarkable performance in vision-language tasks, but their application in the medical field remains underexplored, particularly for integrating structured time series data with unstructured clinical notes. In clinical practice, dynamic time series data such as lab test results capture critical temporal patterns, while clinical notes provide rich semantic context. Merging these modalities is challenging due to the inherent differences between continuous signals and discrete text. To bridge this gap, we introduce ProMedTS, a novel self-supervised multimodal framework that employs prompt-guided learning to unify these heterogeneous data types. Our approach leverages lightweight anomaly detection to generate anomaly captions that serve as prompts, guiding the encoding of raw time series data into informative embeddings. These embeddings are aligned with textual representations in a shared latent space, preserving fine-grained temporal nuances alongside semantic insights. Furthermore, our framework incorporates tailored self-supervised objectives to enhance both intra- and inter-modal alignment. We evaluate ProMedTS on disease diagnosis tasks using real-world datasets, and the results demonstrate that our method consistently outperforms state-of-the-art approaches.
Authors: Yigit Korkmaz, Erdem B{\i}y{\i}k
Abstract: Imitation learning techniques have been shown to be highly effective in real-world control scenarios, such as robotics. However, these approaches not only suffer from compounding error issues but also require human experts to provide complete trajectories. Although there exist interactive methods where an expert oversees the robot and intervenes if needed, these extensions usually only utilize the data collected during intervention periods and ignore the feedback signal hidden in non-intervention timesteps. In this work, we create a model to formulate how the interventions occur in such cases, and show that it is possible to learn a policy with just a handful of expert interventions. Our key insight is that it is possible to get crucial information about the quality of the current state and the optimality of the chosen action from expert feedback, regardless of the presence or the absence of intervention. We evaluate our method on various discrete and continuous simulation environments, a real-world robotic manipulation task, as well as a human subject study. Videos and the code can be found at https://liralab.usc.edu/mile .
Authors: Wei Dai, Steven Wang, Jun Liu
Abstract: Efficient evaluation of three-dimensional (3D) medical images is crucial for diagnostic and therapeutic practices in healthcare. Recent years have seen a substantial uptake in applying deep learning and computer vision to analyse and interpret medical images. Traditional approaches, such as convolutional neural networks (CNNs) and vision transformers (ViTs), face significant computational challenges, prompting the need for architectural advancements. Recent efforts have led to the introduction of novel architectures like the ``Mamba'' model as alternative solutions to traditional CNNs or ViTs. The Mamba model excels in the linear processing of one-dimensional data with low computational demands. However, Mamba's potential for 3D medical image analysis remains underexplored and could face significant computational challenges as the dimension increases. This manuscript presents MobileViM, a streamlined architecture for efficient segmentation of 3D medical images. In the MobileViM network, we invent a new dimension-independent mechanism and a dual-direction traversing approach to incorporate with a vision-Mamba-based framework. MobileViM also features a cross-scale bridging technique to improve efficiency and accuracy across various medical imaging modalities. With these enhancements, MobileViM achieves segmentation speeds exceeding 90 frames per second (FPS) on a single graphics processing unit (i.e., NVIDIA RTX 4090). This performance is over 24 FPS faster than the state-of-the-art deep learning models for processing 3D images with the same computational resources. In addition, experimental evaluations demonstrate that MobileViM delivers superior performance, with Dice similarity scores reaching 92.72%, 86.69%, 80.46%, and 77.43% for PENGWIN, BraTS2024, ATLAS, and Toothfairy2 datasets, respectively, which significantly surpasses existing models.
Authors: Yanzeng Li, Yunfan Xiong, Jialun Zhong, Jinchao Zhang, Jie Zhou, Lei Zou
Abstract: The rise of Large Language Models (LLMs) has led to significant applications but also introduced serious security threats, particularly from jailbreak attacks that manipulate output generation. These attacks utilize prompt engineering and logit manipulation to steer models toward harmful content, prompting LLM providers to implement filtering and safety alignment strategies. We investigate LLMs' safety mechanisms and their recent applications, revealing a new threat model targeting structured output interfaces, which enable attackers to manipulate the inner logit during LLM generation, requiring only API access permissions. To demonstrate this threat model, we introduce a black-box attack framework called AttackPrefixTree (APT). APT exploits structured output interfaces to dynamically construct attack patterns. By leveraging prefixes of models' safety refusal response and latent harmful outputs, APT effectively bypasses safety measures. Experiments on benchmark datasets indicate that this approach achieves higher attack success rate than existing methods. This work highlights the urgent need for LLM providers to enhance security protocols to address vulnerabilities arising from the interaction between safety patterns and structured outputs.
Authors: Jun Zhang, Jue Wang, Huan Li, Lidan Shou, Ke Chen, Yang You, Guiming Xie, Xuejian Gong, Kunlong Zhou
Abstract: Large Language Models (LLMs) have significantly advanced natural language processing with exceptional task generalization capabilities. Low-Rank Adaption (LoRA) offers a cost-effective fine-tuning solution, freezing the original model parameters and training only lightweight, low-rank adapter matrices. However, the memory footprint of LoRA is largely dominated by the original model parameters. To mitigate this, we propose LoRAM, a memory-efficient LoRA training scheme founded on the intuition that many neurons in over-parameterized LLMs have low training utility but are essential for inference. LoRAM presents a unique twist: it trains on a pruned (small) model to obtain pruned low-rank matrices, which are then recovered and utilized with the original (large) model for inference. Additionally, minimal-cost continual pre-training, performed by the model publishers in advance, aligns the knowledge discrepancy between pruned and original models. Our extensive experiments demonstrate the efficacy of LoRAM across various pruning strategies and downstream tasks. For a model with 70 billion parameters, LoRAM enables training on a GPU with only 20G HBM, replacing an A100-80G GPU for LoRA training and 15 GPUs for full fine-tuning. Specifically, QLoRAM implemented by structured pruning combined with 4-bit quantization, for LLaMA-3.1-70B (LLaMA-2-70B), reduces the parameter storage cost that dominates the memory usage in low-rank matrix training by 15.81$\times$ (16.95$\times$), while achieving dominant performance gains over both the original LLaMA-3.1-70B (LLaMA-2-70B) and LoRA-trained LLaMA-3.1-8B (LLaMA-2-13B).
Authors: Guang Ping He
Abstract: The Harrow-Hassidim-Lloyd (HHL) algorithm offers exponential speedup for solving the quantum linear-system problem. But some caveats for the speedup could be hard to met. One of the difficulties is the encoding bottleneck, i.e., the efficient preparation of the initial quantum state. To prepare an arbitrary $N$-dimensional state exactly, existing state-preparation approaches generally require a runtime of $O(N)$, which will ruin the speedup of the HHL algorithm. Here we show that the states can be prepared approximately with a runtime of $O(poly(\log N))$ by employing a slightly modified version of the HHL algorithm itself. Thus, applying this approach to prepare the initial state of the original HHL algorithm can preserve the exponential speedup advantage. It can also serve as a standalone solution for other applications demanding rapid state preparation.
Authors: Qingfa Xiao, Jiachuan Wang, Haoyang Li, Cheng Deng, Jiaqi Tang, Shuangyin Li, Yongqi Zhang, Jun Wang, Lei Chen
Abstract: Recent advances in large language models (LLMs) have showcased exceptional performance in long-context tasks, while facing significant inference efficiency challenges with limited GPU memory. Existing solutions first proposed the sliding-window approach to accumulate a set of historical \textbf{key-value} (KV) pairs for reuse, then further improvements selectively retain its subsets at each step. However, due to the sparse attention distribution across a long context, it is hard to identify and recall relevant KV pairs, as the attention is distracted by massive candidate pairs. Additionally, we found it promising to select representative tokens as probe-Query in each sliding window to effectively represent the entire context, which is an approach overlooked by existing methods. Thus, we propose \textbf{ActQKV}, a training-free, \textbf{Act}ivation-aware approach that dynamically determines probe-\textbf{Q}uery and leverages it to retrieve the relevant \textbf{KV} pairs for inference. Specifically, ActQKV monitors a token-level indicator, Activation Bias, within each context window, enabling the proper construction of probe-Query for retrieval at pre-filling stage. To accurately recall the relevant KV pairs and minimize the irrelevant ones, we design a dynamic KV cut-off mechanism guided by information density across layers at the decoding stage. Experiments on the Long-Bench and $\infty$ Benchmarks demonstrate its state-of-the-art performance with competitive inference quality and resource efficiency.
Authors: Peiwen Yuan, Chuyi Tan, Shaoxiong Feng, Yiwei Li, Xinglin Wang, Yueqi Zhang, Jiayi Shi, Boyuan Pan, Yao Hu, Kan Li
Abstract: Despite the rapid progress of large language models (LLMs), their length-controllable text generation (LCTG) ability remains below expectations, posing a major limitation for practical applications. Existing methods mainly focus on end-to-end training to reinforce adherence to length constraints. However, the lack of decomposition and targeted enhancement of LCTG sub-abilities restricts further progress.To bridge this gap, we conduct a bottom-up decomposition of LCTG sub-abilities with human patterns as reference and perform a detailed error analysis.On this basis, we propose MarkerGen, a simple-yet-effective plug-and-play approach that:(1) mitigates LLM fundamental deficiencies via external tool integration;(2) conducts explicit length modeling with dynamically inserted markers;(3) employs a three-stage generation scheme to better align length constraints while maintaining content quality.Comprehensive experiments demonstrate that MarkerGen significantly improves LCTG across various settings, exhibiting outstanding effectiveness and generalizability.
Authors: Yushi Feng, Tsai Hor Chan, Guosheng Yin, Lequan Yu
Abstract: Data augmentation is necessary for graph representation learning due to the scarcity and noise present in graph data. Most of the existing augmentation methods overlook the context information inherited from the dataset as they rely solely on the graph structure for augmentation. Despite the success of some large language model-based (LLM) graph learning methods, they are mostly white-box which require access to the weights or latent features from the open-access LLMs, making them difficult to be democratized for everyone as existing LLMs are mostly closed-source for commercial considerations. To overcome these limitations, we propose a black-box context-driven graph data augmentation approach, with the guidance of LLMs -- DemoGraph. Leveraging the text prompt as context-related information, we task the LLM with generating knowledge graphs (KGs), which allow us to capture the structural interactions from the text outputs. We then design a dynamic merging schema to stochastically integrate the LLM-generated KGs into the original graph during training. To control the sparsity of the augmented graph, we further devise a granularity-aware prompting strategy and an instruction fine-tuning module, which seamlessly generates text prompts according to different granularity levels of the dataset. Extensive experiments on various graph learning tasks validate the effectiveness of our method over existing graph data augmentation methods. Notably, our approach excels in scenarios involving electronic health records (EHRs), which validates its maximal utilization of contextual knowledge, leading to enhanced predictive performance and interpretability.
Authors: Jintang Li, Ruofan Wu, Yuchang Zhu, Huizhe Zhang, Liang Chen, Zibin Zheng
Abstract: Large language models (LLMs) have demonstrated remarkable in-context reasoning capabilities across a wide range of tasks, particularly with unstructured inputs such as language or images. However, LLMs struggle to handle structured data, such as graphs, due to their lack of understanding of non-Euclidean structures. As a result, without additional fine-tuning, their performance significantly lags behind that of graph neural networks (GNNs) in graph learning tasks. In this paper, we show that learning on graph data can be conceptualized as a retrieval-augmented generation (RAG) process, where specific instances (e.g., nodes or edges) act as queries, and the graph itself serves as the retrieved context. Building on this insight, we propose a series of RAG frameworks to enhance the in-context learning capabilities of LLMs for graph learning tasks. Comprehensive evaluations demonstrate that our proposed RAG frameworks significantly improve LLM performance on graph-based tasks, particularly in scenarios where a pretrained LLM must be used without modification or accessed via an API.
Authors: Peiwen Yuan, Yueqi Zhang, Shaoxiong Feng, Yiwei Li, Xinglin Wang, Jiayi Shi, Chuyi Tan, Boyuan Pan, Yao Hu, Kan Li
Abstract: Evaluating models on large benchmarks is very resource-intensive, especially during the period of rapid model evolution. Existing efficient evaluation methods estimate the performance of target models by testing them only on a small and static coreset of the benchmark, which is derived from the publicly available evaluation results of source models. These methods rely on the assumption that target models have high prediction consistency with source models. However, we demonstrate that it doesn't generalize well in practice. To alleviate the inconsistency issue, we present TailoredBench, a method that conducts customized evaluation tailored to each target model. Specifically, a Global-coreset is first constructed as a probe to identify the most consistent source models for each target model with an adaptive source model selection strategy. Afterwards, a scalable K-Medoids clustering algorithm is proposed to extend the Global-coreset to a tailored Native-coreset for each target model. According to the predictions on Native-coresets, we obtain the performance of target models on the whole benchmark with a calibrated estimation strategy. Comprehensive experiments on 5 benchmarks across over 300 models demonstrate that compared to best performing baselines, TailoredBench achieves an average reduction of 31.4% in MAE of accuracy estimates under the same inference budgets, showcasing strong effectiveness and generalizability.
Authors: Kenneth Enevoldsen, Isaac Chung, Imene Kerboua, M\'arton Kardos, Ashwin Mathur, David Stap, Jay Gala, Wissam Siblini, Dominik Krzemi\'nski, Genta Indra Winata, Saba Sturua, Saiteja Utpala, Mathieu Ciancone, Marion Schaeffer, Gabriel Sequeira, Diganta Misra, Shreeya Dhakal, Jonathan Rystr{\o}m, Roman Solomatin, \"Omer \c{C}a\u{g}atan, Akash Kundu, Martin Bernstorff, Shitao Xiao, Akshita Sukhlecha, Bhavish Pahwa, Rafa{\l} Po\'swiata, Kranthi Kiran GV, Shawon Ashraf, Daniel Auras, Bj\"orn Pl\"uster, Jan Philipp Harries, Lo\"ic Magne, Isabelle Mohr, Mariya Hendriksen, Dawei Zhu, Hippolyte Gisserot-Boukhlef, Tom Aarsen, Jan Kostkan, Konrad Wojtasik, Taemin Lee, Marek \v{S}uppa, Crystina Zhang, Roberta Rocca, Mohammed Hamdy, Andrianos Michail, John Yang, Manuel Faysse, Aleksei Vatolin, Nandan Thakur, Manan Dey, Dipam Vasani, Pranjal Chitale, Simone Tedeschi, Nguyen Tai, Artem Snegirev, Michael G\"unther, Mengzhou Xia, Weijia Shi, Xing Han L\`u, Jordan Clive, Gayatri Krishnakumar, Anna Maksimova, Silvan Wehrli, Maria Tikhonova, Henil Panchal, Aleksandr Abramov, Malte Ostendorff, Zheng Liu, Simon Clematide, Lester James Miranda, Alena Fenogenova, Guangyu Song, Ruqiya Bin Safi, Wen-Ding Li, Alessia Borghini, Federico Cassano, Hongjin Su, Jimmy Lin, Howard Yen, Lasse Hansen, Sara Hooker, Chenghao Xiao, Vaibhav Adlakha, Orion Weller, Siva Reddy, Niklas Muennighoff
Abstract: Text embeddings are typically evaluated on a limited set of tasks, which are constrained by language, domain, and task diversity. To address these limitations and provide a more comprehensive evaluation, we introduce the Massive Multilingual Text Embedding Benchmark (MMTEB) - a large-scale, community-driven expansion of MTEB, covering over 500 quality-controlled evaluation tasks across 250+ languages. MMTEB includes a diverse set of challenging, novel tasks such as instruction following, long-document retrieval, and code retrieval, representing the largest multilingual collection of evaluation tasks for embedding models to date. Using this collection, we develop several highly multilingual benchmarks, which we use to evaluate a representative set of models. We find that while large language models (LLMs) with billions of parameters can achieve state-of-the-art performance on certain language subsets and task categories, the best-performing publicly available model is multilingual-e5-large-instruct with only 560 million parameters. To facilitate accessibility and reduce computational cost, we introduce a novel downsampling method based on inter-task correlation, ensuring a diverse selection while preserving relative model rankings. Furthermore, we optimize tasks such as retrieval by sampling hard negatives, creating smaller but effective splits. These optimizations allow us to introduce benchmarks that drastically reduce computational demands. For instance, our newly introduced zero-shot English benchmark maintains a ranking order similar to the full-scale version but at a fraction of the computational cost.
Authors: Dario Garcia-Gasulla, Anna Arias-Duart, Adrian Tormos, Daniel Hinjos, Oscar Molina-Sedano, Ashwin Kumar Gururajan, Maria Eugenia Cardello
Abstract: Direct Preference Optimization (DPO) is an efficient alignment technique that steers LLMs towards preferable outputs by training on preference data, bypassing the need for explicit reward models. Its simplicity enables easy adaptation to various domains and safety requirements. This paper examines DPO's effectiveness in model safety against jailbreaking attacks while minimizing data requirements and training costs. We introduce Egida, a dataset expanded from multiple sources, which includes 27 different safety topics and 18 different attack styles, complemented with synthetic and human labels. This data is used to boost the safety of state-of-the-art LLMs (Llama-3.1-8B/70B-Instruct, Qwen-2.5-7B/72B-Instruct) across topics and attack styles. In addition to safety evaluations, we assess their post-alignment performance degradation in general purpose tasks, and their tendency to over refusal. Following the proposed methodology, trained models reduce their Attack Success Rate by 10%-30%, using small training efforts (2,000 samples) with low computational cost (3\$ for 8B models, 20\$ for 72B models). Safety aligned models generalize to unseen topics and attack styles, with the most successful attack style reaching a success rate around 5%. Size and family are found to strongly influence model malleability towards safety, pointing at the importance of pre-training choices. To validate our findings, a large independent assessment of human preference agreement with Llama-Guard-3-8B is conducted by the authors and the associated dataset Egida-HSafe is released. Overall, this study illustrates how affordable and accessible it is to enhance LLM safety using DPO while outlining its current limitations. All datasets and models are released to enable reproducibility and further research.
Authors: Takuya Matsuyama, Shinji Nishimoto, Yu Takagi
Abstract: Understanding the property of neural populations (or voxels) in the human brain can advance our comprehension of human perceptual and cognitive processing capabilities and contribute to developing brain-inspired computer models. Recent encoding models using deep neural networks (DNNs) have successfully predicted voxel-wise activity. However, interpreting the properties that explain voxel responses remains challenging because of the black-box nature of DNNs. As a solution, we propose LLM-assisted Visual Cortex Captioning (LaVCa), a data-driven approach that uses large language models (LLMs) to generate natural-language captions for images to which voxels are selective. By applying LaVCa for image-evoked brain activity, we demonstrate that LaVCa generates captions that describe voxel selectivity more accurately than the previously proposed method. Furthermore, the captions generated by LaVCa quantitatively capture more detailed properties than the existing method at both the inter-voxel and intra-voxel levels. Furthermore, a more detailed analysis of the voxel-specific properties generated by LaVCa reveals fine-grained functional differentiation within regions of interest (ROIs) in the visual cortex and voxels that simultaneously represent multiple distinct concepts. These findings offer profound insights into human visual representations by assigning detailed captions throughout the visual cortex while highlighting the potential of LLM-based methods in understanding brain representations. Please check out our webpage at https://sites.google.com/view/lavca-llm/
Authors: Guilherme Sousa, Rinaldo Lima, Cassia Trojahn
Abstract: Ontology, and more broadly, Knowledge Graph Matching is a challenging task in which expressiveness has not been fully addressed. Despite the increasing use of embeddings and language models for this task, approaches for generating expressive correspondences still do not take full advantage of these models, in particular, large language models (LLMs). This paper proposes to integrate LLMs into an approach for generating expressive correspondences based on alignment need and ABox-based relation discovery. The generation of correspondences is performed by matching similar surroundings of instance sub-graphs. The integration of LLMs results in different architectural modifications, including label similarity, sub-graph matching, and entity matching. The performance word embeddings, sentence embeddings, and LLM-based embeddings, was compared. The results demonstrate that integrating LLMs surpasses all other models, enhancing the baseline version of the approach with a 45\% increase in F-measure.
Authors: Francesco Pontiggia, Filip Mac\'ak, Roman Andriushchenko, Michele Chiari, Milan \v{C}e\v{s}ka
Abstract: Multi-agent planning under stochastic dynamics is usually formalised using decentralized (partially observable) Markov decision processes ( MDPs) and reachability or expected reward specifications. In this paper, we propose a different approach: we use an MDP describing how a single agent operates in an environment and probabilistic hyperproperties to capture desired temporal objectives for a set of decentralized agents operating in the environment. We extend existing approaches for model checking probabilistic hyperproperties to handle temporal formulae relating paths of different agents, thus requiring the self-composition between multiple MDPs. Using several case studies, we demonstrate that our approach provides a flexible and expressive framework to broaden the specification capabilities with respect to existing planning techniques. Additionally, we establish a close connection between a subclass of probabilistic hyperproperties and planning for a particular type of Dec-MDPs, for both of which we show undecidability. This lays the ground for the use of existing decentralized planning tools in the field of probabilistic hyperproperty verification.
Authors: DongGeon Lee, Hwanjo Yu
Abstract: Hallucinations in large language model (LLM) outputs severely limit their reliability in knowledge-intensive tasks such as question answering. To address this challenge, we introduce REFIND (Retrieval-augmented Factuality hallucINation Detection), a novel framework that detects hallucinated spans within LLM outputs by directly leveraging retrieved documents. As part of the REFIND, we propose the Context Sensitivity Ratio (CSR), a novel metric that quantifies the sensitivity of LLM outputs to retrieved evidence. This innovative approach enables REFIND to efficiently and accurately detect hallucinations, setting it apart from existing methods. In the evaluation, REFIND demonstrated robustness across nine languages, including low-resource settings, and significantly outperformed baseline models, achieving superior IoU scores in identifying hallucinated spans. This work highlights the effectiveness of quantifying context sensitivity for hallucination detection, thereby paving the way for more reliable and trustworthy LLM applications across diverse languages.
Authors: Or Raphael Bidusa, Shaul Markovitch
Abstract: The opaque nature of Large Language Models (LLMs) has led to significant research efforts aimed at enhancing their interpretability, primarily through post-hoc methods. More recent in-hoc approaches, such as Concept Bottleneck Models (CBMs), offer both interpretability and intervenability by incorporating explicit concept representations. However, these methods suffer from key limitations, including reliance on labeled concept datasets and significant architectural modifications that challenges re-integration into existing system pipelines. In this work, we introduce a new methodology for incorporating interpretability and intervenability into an existing model by integrating Concept Layers (CLs) into its architecture. Our approach projects the model's internal vector representations into a conceptual, explainable vector space before reconstructing and feeding them back into the model. Furthermore, we eliminate the need for a human-selected concept set by algorithmically searching an ontology for a set of concepts that can be either task-specific or task-agnostic. We evaluate CLs across multiple tasks, demonstrating that they maintain the original model's performance and agreement while enabling meaningful interventions. Additionally, we present a proof of concept showcasing an intervenability interface, allowing users to adjust model behavior dynamically, such as mitigating biases during inference.
Authors: Julian Vexler, Bj\"orn Vieten, Martin Nelke, Stefan Kramer
Abstract: We present CavePerception, a framework for the analysis of sparse data from sensor networks that incorporates elements of inverse modeling and forward modeling. By integrating machine learning with physical modeling in a hypotheses space, we aim to improve the interpretability of sparse, noisy, and potentially incomplete sensor data. The framework assumes data from a two-dimensional sensor network laid out in a graph structure that detects certain objects, with certain motion patterns. Examples of such sensors are magnetometers. Given knowledge about the objects and the way they act on the sensors, one can develop a data generator that produces data from simulated motions of the objects across the sensor field. The framework uses the simulated data to infer object behaviors across the sensor network. The approach is experimentally tested on real-world data, where magnetometers are used on an airport to detect and identify aircraft motions. Experiments demonstrate the value of integrating inverse and forward modeling, enabling intelligent systems to better understand and predict complex, sensor-driven events.
Authors: Feiye Huo, Jianchao Tan, Kefeng Zhang, Xunliang Cai, Shengli Sun
Abstract: The growing scale of Large Language Models (LLMs) has exacerbated inference latency and computational costs. Speculative decoding methods, which aim to mitigate these issues, often face inefficiencies in the construction of token trees and the verification of candidate tokens. Existing strategies, including chain mode, static tree, and dynamic tree approaches, have limitations in accurately preparing candidate token trees for verification. We propose a novel method named C2T that adopts a lightweight classifier to generate and prune token trees dynamically. Our classifier considers additional feature variables beyond the commonly used joint probability to predict the confidence score for each draft token to determine whether it is the candidate token for verification. This method outperforms state-of-the-art (SOTA) methods such as EAGLE-2 on multiple benchmarks, by reducing the total number of candidate tokens by 25% while maintaining or even improving the acceptance length.
Authors: Tim Baumg\"artner, Ted Briscoe, Iryna Gurevych
Abstract: We present PeerQA, a real-world, scientific, document-level Question Answering (QA) dataset. PeerQA questions have been sourced from peer reviews, which contain questions that reviewers raised while thoroughly examining the scientific article. Answers have been annotated by the original authors of each paper. The dataset contains 579 QA pairs from 208 academic articles, with a majority from ML and NLP, as well as a subset of other scientific communities like Geoscience and Public Health. PeerQA supports three critical tasks for developing practical QA systems: Evidence retrieval, unanswerable question classification, and answer generation. We provide a detailed analysis of the collected dataset and conduct experiments establishing baseline systems for all three tasks. Our experiments and analyses reveal the need for decontextualization in document-level retrieval, where we find that even simple decontextualization approaches consistently improve retrieval performance across architectures. On answer generation, PeerQA serves as a challenging benchmark for long-context modeling, as the papers have an average size of 12k tokens. Our code and data is available at https://github.com/UKPLab/peerqa.
Authors: Ruida Hu, Chao Peng, Xinchen Wang, Cuiyun Gao
Abstract: Environment configuration is a critical yet time-consuming step in software development, especially when dealing with unfamiliar code repositories. While Large Language Models (LLMs) demonstrate the potential to accomplish software engineering tasks, existing methods for environment configuration often rely on manual efforts or fragile scripts, leading to inefficiencies and unreliable outcomes. We introduce Repo2Run, the first LLM-based agent designed to fully automate environment configuration and generate executable Dockerfiles for arbitrary Python repositories. We address two major challenges: (1) enabling the LLM agent to configure environments within isolated Docker containers, and (2) ensuring the successful configuration process is recorded and accurately transferred to a Dockerfile without error. To achieve this, we propose atomic configuration synthesis, featuring a dual-environment architecture (internal and external environment) with a rollback mechanism to prevent environment "pollution" from failed commands, guaranteeing atomic execution (execute fully or not at all) and a Dockerfile generator to transfer successful configuration steps into runnable Dockerfiles. We evaluate Repo2Run~on our proposed benchmark of 420 recent Python repositories with unit tests, where it achieves an 86.0% success rate, outperforming the best baseline by 63.9%.
Authors: Jusen Du, Weigao Sun, Disen Lan, Jiaxi Hu, Yu Cheng
Abstract: Linear sequence modeling methods, such as linear attention, state space modeling, and linear RNNs, offer significant efficiency improvements by reducing the complexity of training and inference. However, these methods typically compress the entire input sequence into a single fixed-size memory state, which leads to suboptimal performance on recall-intensive downstream tasks. Drawing inspiration from neuroscience, particularly the brain's ability to maintain robust long-term memory while mitigating "memory interference", we introduce a novel architecture called Mixture-of-Memories (MoM). MoM utilizes multiple independent memory states, with a router network directing input tokens to specific memory states. This approach greatly enhances the overall memory capacity while minimizing memory interference. As a result, MoM performs exceptionally well on recall-intensive tasks, surpassing existing linear sequence modeling techniques. Despite incorporating multiple memory states, the computation of each memory state remains linear in complexity, allowing MoM to retain the linear-complexity advantage during training, while constant-complexity during inference. Our experimental results show that MoM significantly outperforms current linear sequence models on downstream language tasks, particularly recall-intensive tasks, and even achieves performance comparable to Transformer models. The code is released at https://github.com/OpenSparseLLMs/MoM and is also released as a part of https://github.com/OpenSparseLLMs/Linear-MoE.
URLs: https://github.com/OpenSparseLLMs/MoM, https://github.com/OpenSparseLLMs/Linear-MoE.
Authors: Yixing Fan, Qiang Yan, Wenshan Wang, Jiafeng Guo, Ruqing Zhang, Xueqi Cheng
Abstract: \Ac{RAG} has emerged as a crucial technique for enhancing large models with real-time and domain-specific knowledge. While numerous improvements and open-source tools have been proposed to refine the \ac{RAG} framework for accuracy, relatively little attention has been given to improving the trustworthiness of generated results. To address this gap, we introduce TrustRAG, a novel framework that enhances \ac{RAG} from three perspectives: indexing, retrieval, and generation. Specifically, in the indexing stage, we propose a semantic-enhanced chunking strategy that incorporates hierarchical indexing to supplement each chunk with contextual information, ensuring semantic completeness. In the retrieval stage, we introduce a utility-based filtering mechanism to identify high-quality information, supporting answer generation while reducing input length. In the generation stage, we propose fine-grained citation enhancement, which detects opinion-bearing sentences in responses and infers citation relationships at the sentence-level, thereby improving citation accuracy. We open-source the TrustRAG framework and provide a demonstration studio designed for excerpt-based question answering tasks \footnote{https://huggingface.co/spaces/golaxy/TrustRAG}. Based on these, we aim to help researchers: 1) systematically enhancing the trustworthiness of \ac{RAG} systems and (2) developing their own \ac{RAG} systems with more reliable outputs.
Authors: Hongbo Zhang, Han Cui, Guangsheng Bao, Linyi Yang, Jun Wang, Yue Zhang
Abstract: We introduce Direct Value Optimization (DVO), an innovative reinforcement learning framework for enhancing large language models in complex reasoning tasks. Unlike traditional methods relying on preference labels, DVO utilizes value signals at individual reasoning steps, optimizing models via a mean squared error loss. The key benefit of DVO lies in its fine-grained supervision, circumventing the need for labor-intensive human annotations. Target values within the DVO are estimated using either Monte Carlo Tree Search or an outcome value model. Our empirical analysis on both mathematical and commonsense reasoning tasks shows that DVO consistently outperforms existing offline preference optimization techniques, even with fewer training steps. These findings underscore the importance of value signals in advancing reasoning capabilities and highlight DVO as a superior methodology under scenarios lacking explicit human preference information.
Authors: Marco Arazzi, Mert Cihangiroglu, Serena Nicolazzo, Antonino Nocera
Abstract: Dataset Distillation (DD) is a powerful technique for reducing large datasets into compact, representative synthetic datasets, accelerating Machine Learning training. However, traditional DD methods operate in a centralized manner, which poses significant privacy threats and reduces its applicability. To mitigate these risks, we propose a Secure Federated Data Distillation framework (SFDD) to decentralize the distillation process while preserving privacy.Unlike existing Federated Distillation techniques that focus on training global models with distilled knowledge, our approach aims to produce a distilled dataset without exposing local contributions. We leverage the gradient-matching-based distillation method, adapting it for a distributed setting where clients contribute to the distillation process without sharing raw data. The central aggregator iteratively refines a synthetic dataset by integrating client-side updates while ensuring data confidentiality. To make our approach resilient to inference attacks perpetrated by the server that could exploit gradient updates to reconstruct private data, we create an optimized Local Differential Privacy approach, called LDPO-RLD (Label Differential Privacy Obfuscation via Randomized Linear Dispersion). Furthermore, we assess the framework's resilience against malicious clients executing backdoor attacks and demonstrate robustness under the assumption of a sufficient number of participating clients. Our experimental results demonstrate the effectiveness of SFDD and that the proposed defense concretely mitigates the identified vulnerabilities, with minimal impact on the performance of the distilled dataset. By addressing the interplay between privacy and federation in dataset distillation, this work advances the field of privacy-preserving Machine Learning making our SFDD framework a viable solution for sensitive data-sharing applications.
Authors: Junqi Jiang, Luca Marzari, Aaryan Purohit, Francesco Leofante
Abstract: The increasing use of Machine Learning (ML) models to aid decision-making in high-stakes industries demands explainability to facilitate trust. Counterfactual Explanations (CEs) are ideally suited for this, as they can offer insights into the predictions of an ML model by illustrating how changes in its input data may lead to different outcomes. However, for CEs to realise their explanatory potential, significant challenges remain in ensuring their robustness under slight changes in the scenario being explained. Despite the widespread recognition of CEs' robustness as a fundamental requirement, a lack of standardised tools and benchmarks hinders a comprehensive and effective comparison of robust CE generation methods. In this paper, we introduce RobustX, an open-source Python library implementing a collection of CE generation and evaluation methods, with a focus on the robustness property. RobustX provides interfaces to several existing methods from the literature, enabling streamlined access to state-of-the-art techniques. The library is also easily extensible, allowing fast prototyping of novel robust CE generation and evaluation methods.
Authors: Ahmad Alomari, Sathish A. P. Kumar
Abstract: This study proposes a GPA for designing optimal Quantum Sensor Circuits (QSCs) to address complex quantum physics problems. The GPA consists of two parts: the Quantum Policy Evaluation (QPE) and the Quantum Policy Improvement (QPI). The QPE performs phase estimation to generate the search space, while the QPI utilizes Grover search and amplitude amplification techniques to efficiently identify an optimal policy that generates optimal QSCs. The GPA generates QSCs by selecting sequences of gates that maximize the Quantum Fisher Information (QFI) while minimizing the number of gates. The QSCs generated by the GPA are capable of producing entangled quantum states, specifically the squeezed states. High QFI indicates increased sensitivity to parameter changes, making the circuit useful for quantum state estimation and control tasks. Evaluation of the GPA on a QSC that consists of two qubits and a sequence of R_x, R_y, and S gates demonstrates its efficiency in generating optimal QSCs with a QFI of 1. Compared to existing quantum agents, the GPA achieves higher QFI with fewer gates, demonstrating a more efficient and scalable approach to the design of QSCs. This work illustrates the potential computational power of quantum agents for solving quantum physics problems
Authors: Wanke Xia, Ruxin Peng, Haoqi Chu, Xinlei Zhu, Zhiyu Yang, Yaojun Wang
Abstract: Rice is one of the most widely cultivated crops globally and has been developed into numerous varieties. The quality of rice during cultivation is primarily determined by its cultivar and characteristics. Traditionally, rice classification and quality assessment rely on manual visual inspection, a process that is both time-consuming and prone to errors. However, with advancements in machine vision technology, automating rice classification and quality evaluation based on its cultivar and characteristics has become increasingly feasible, enhancing both accuracy and efficiency. This study proposes a real-time evaluation mechanism for comprehensive rice grain assessment, integrating a one-stage object detection approach, a deep convolutional neural network, and traditional machine learning techniques. The proposed framework enables rice variety identification, grain completeness grading, and grain chalkiness evaluation. The rice grain dataset used in this study comprises approximately 20,000 images from six widely cultivated rice varieties in China. Experimental results demonstrate that the proposed mechanism achieves a mean average precision (mAP) of 99.14% in the object detection task and an accuracy of 97.89% in the classification task. Furthermore, the framework attains an average accuracy of 97.56% in grain completeness grading within the same rice variety, contributing to an effective quality evaluation system.
Authors: Abhik Roychoudhury, Corina Pasareanu, Michael Pradel, Baishakhi Ray
Abstract: Large Language Models (LLMs) have shown surprising proficiency in generating code snippets, promising to automate large parts of software engineering via artificial intelligence (AI). We argue that successfully deploying AI software engineers requires a level of trust equal to or even greater than the trust established by human-driven software engineering practices. The recent trend toward LLM agents offers a path toward integrating the power of LLMs to create new code with the power of analysis tools to increase trust in the code. This opinion piece comments on whether LLM agents could dominate software engineering workflows in the future and whether the focus of programming will shift from programming at scale to programming with trust.
Authors: Anudeex Shetty, Amin Beheshti, Mark Dras, Usman Naseem
Abstract: Alignment techniques have become central to ensuring that Large Language Models (LLMs) generate outputs consistent with human values. However, existing alignment paradigms often model an averaged or monolithic preference, failing to account for the diversity of perspectives across cultures, demographics, and communities. This limitation is particularly critical in health-related scenarios, where plurality is essential due to the influence of culture, religion, personal values, and conflicting opinions. Despite progress in pluralistic alignment, no prior work has focused on health, likely due to the unavailability of publicly available datasets. To address this gap, we introduce VITAL, a new benchmark dataset comprising 13.1K value-laden situations and 5.4K multiple-choice questions focused on health, designed to assess and benchmark pluralistic alignment methodologies. Through extensive evaluation of eight LLMs of varying sizes, we demonstrate that existing pluralistic alignment techniques fall short in effectively accommodating diverse healthcare beliefs, underscoring the need for tailored AI alignment in specific domains. This work highlights the limitations of current approaches and lays the groundwork for developing health-specific alignment solutions.
Authors: Jiaqi Li, Xizhong Guo, Yang Zhao, Lvyang Zhang, Lidong Zhai
Abstract: Rapid industrial digitalization has created intricate cybersecurity demands that necessitate effective validation methods. While cyber ranges and simulation platforms are widely deployed, they frequently face limitations in scenario diversity and creation efficiency. In this paper, we present SpiderSim, a theoretical cybersecurity simulation platform enabling rapid and lightweight scenario generation for industrial digitalization security research. At its core, our platform introduces three key innovations: a structured framework for unified scenario modeling, a multi-agent collaboration mechanism for automated generation, and modular atomic security capabilities for flexible scenario composition. Extensive implementation trials across multiple industrial digitalization contexts, including marine ranch monitoring systems, validate our platform's capacity for broad scenario coverage with efficient generation processes. Built on solid theoretical foundations and released as open-source software, SpiderSim facilitates broader research and development in automated security testing for industrial digitalization.
Authors: Matthew Wood, Mathieu Klop, Maxime Allard
Abstract: mRNA-based vaccines have become a major focus in the pharmaceutical industry. The coding sequence as well as the Untranslated Regions (UTRs) of an mRNA can strongly influence translation efficiency, stability, degradation, and other factors that collectively determine a vaccine's effectiveness. However, optimizing mRNA sequences for those properties remains a complex challenge. Existing deep learning models often focus solely on coding region optimization, overlooking the UTRs. We present Helix-mRNA, a structured state-space-based and attention hybrid model to address these challenges. In addition to a first pre-training, a second pre-training stage allows us to specialise the model with high-quality data. We employ single nucleotide tokenization of mRNA sequences with codon separation, ensuring prior biological and structural information from the original mRNA sequence is not lost. Our model, Helix-mRNA, outperforms existing methods in analysing both UTRs and coding region properties. It can process sequences 6x longer than current approaches while using only 10% of the parameters of existing foundation models. Its predictive capabilities extend to all mRNA regions. We open-source the model (https://github.com/helicalAI/helical) and model weights (https://huggingface.co/helical-ai/helix-mRNA).
URLs: https://github.com/helicalAI/helical), https://huggingface.co/helical-ai/helix-mRNA).
Authors: Yifei Yang, Zouying Cao, Xinbei Ma, Yao Yao, Libo Qin, Zhi Chen, Hai Zhao
Abstract: Training Large Language Models (LLMs) from scratch requires immense computational resources, making it prohibitively expensive. Model scaling-up offers a promising solution by leveraging the parameters of smaller models to create larger ones. However, existing depth scaling-up methods rely on empirical heuristic rules for layer duplication, which result in poorer initialization and slower convergence during continual pre-training. We propose \textbf{LESA}, a novel learnable method for depth scaling-up. By concatenating parameters from each layer and applying Singular Value Decomposition, we uncover latent patterns between layers, suggesting that inter-layer parameters can be learned. LESA uses a neural network to predict the parameters inserted between adjacent layers, enabling better initialization and faster training. Experiments show that LESA outperforms existing baselines, achieving superior performance with less than half the computational cost during continual pre-training. Extensive analyses demonstrate its effectiveness across different model sizes and tasks.
Authors: Tianqing Wang, Xun Xue, Guoliang Li, Yong Wang
Abstract: In this demonstration, we present AnDB, an AI-native database that supports traditional OLTP workloads and innovative AI-driven tasks, enabling unified semantic analysis across structured and unstructured data. While structured data analytics is mature, challenges remain in bridging the semantic gap between user queries and unstructured data. AnDB addresses these issues by leveraging cutting-edge AI-native technologies, allowing users to perform semantic queries using intuitive SQL-like statements without requiring AI expertise. This approach eliminates the ambiguity of traditional text-to-SQL systems and provides a seamless end-to-end optimization for analyzing all data types. AnDB automates query processing by generating multiple execution plans and selecting the optimal one through its optimizer, which balances accuracy, execution time, and financial cost based on user policies and internal optimizing mechanisms. AnDB future-proofs data management infrastructure, empowering users to effectively and efficiently harness the full potential of all kinds of data without starting from scratch.
Authors: Peter Carragher, Abhinand Jha, R Raghav, Kathleen M. Carley
Abstract: Large Language Models (LLMs) demonstrate remarkable capabilities in question answering (QA), but metrics for assessing their reliance on memorization versus retrieval remain underdeveloped. Moreover, while finetuned models are state-of-the-art on closed-domain tasks, general-purpose models like GPT-4o exhibit strong zero-shot performance. This raises questions about the trade-offs between memorization, generalization, and retrieval. In this work, we analyze the extent to which multimodal retrieval-augmented VLMs memorize training data compared to baseline VLMs. Using the WebQA benchmark, we contrast finetuned models with baseline VLMs on multihop retrieval and question answering, examining the impact of finetuning on data memorization. To quantify memorization in end-to-end retrieval and QA systems, we propose several proxy metrics by investigating instances where QA succeeds despite retrieval failing. Our results reveal the extent to which finetuned models rely on memorization. In contrast, retrieval-augmented VLMs have lower memorization scores, at the cost of accuracy (72% vs 52% on WebQA test set). As such, our measures pose a challenge for future work to reconcile memorization and generalization in both Open-Domain QA and joint Retrieval-QA tasks.
Authors: Jiahao Liu, Dongsheng Li, Hansu Gu, Peng Zhang, Tun Lu, Li Shang, Ning Gu
Abstract: Recommender systems often suffer from popularity bias, where frequently interacted items are overrepresented in recommendations. This bias stems from propensity factors influencing training data, leading to imbalanced exposure. In this paper, we introduce a Fair Sampling (FS) approach to address this issue by ensuring that both users and items are selected with equal probability as positive and negative instances. Unlike traditional inverse propensity score (IPS) methods, FS does not require propensity estimation, eliminating errors associated with inaccurate calculations. Our theoretical analysis demonstrates that FS effectively neutralizes the influence of propensity factors, achieving unbiased learning. Experimental results validate that FS outperforms state-of-the-art methods in both point-wise and pair-wise recommendation tasks, enhancing recommendation fairness without sacrificing accuracy. The implementation is available at https://anonymous.4open.science/r/Fair-Sampling.
Authors: Jiahao Liu, Shengkang Gu, Dongsheng Li, Guangping Zhang, Mingzhe Han, Hansu Gu, Peng Zhang, Tun Lu, Li Shang, Ning Gu
Abstract: Large Language Model (LLM)-based user agents have emerged as a powerful tool for improving recommender systems by simulating user interactions. However, existing methods struggle with cross-domain scenarios due to inefficient memory structures, leading to irrelevant information retention and failure to account for social influence factors such as popularity. To address these limitations, we introduce AgentCF++, a novel framework featuring a dual-layer memory architecture and a two-step fusion mechanism to filter domain-specific preferences effectively. Additionally, we propose interest groups with shared memory, allowing the model to capture the impact of popularity trends on users with similar interests. Through extensive experiments on multiple cross-domain datasets, AgentCF++ demonstrates superior performance over baseline models, highlighting its effectiveness in refining user behavior simulation for recommender systems. Our code is available at https://anonymous.4open.science/r/AgentCF-plus.
Authors: Jiahao Liu, Xueshuo Yan, Dongsheng Li, Guangping Zhang, Hansu Gu, Peng Zhang, Tun Lu, Li Shang, Ning Gu
Abstract: Current recommendation systems powered by large language models (LLMs) often underutilize their reasoning capabilities due to a lack of explicit logical structuring. To address this limitation, we introduce CoT-Rec, a framework that integrates Chain-of-Thought (CoT) reasoning into LLM-driven recommendations by incorporating two crucial processes: user preference analysis and item perception evaluation. CoT-Rec operates in two key phases: (1) personalized data extraction, where user preferences and item perceptions are identified, and (2) personalized data application, where this information is leveraged to refine recommendations. Our experimental analysis demonstrates that CoT-Rec improves recommendation accuracy by making better use of LLMs' reasoning potential. The implementation is publicly available at https://anonymous.4open.science/r/CoT-Rec.
Authors: Feiyuan Zhang, Dezhi Zhu, James Ming, Yilun Jin, Di Chai, Liu Yang, Han Tian, Zhaoxin Fan, Kai Chen
Abstract: Retrieval-Augmented Generation (RAG) systems have shown substantial benefits in applications such as question answering and multi-turn dialogue \citep{lewis2020retrieval}. However, traditional RAG methods, while leveraging static knowledge bases, often overlook the potential of dynamic historical information in ongoing conversations. To bridge this gap, we introduce DH-RAG, a Dynamic Historical Context-Powered Retrieval-Augmented Generation Method for Multi-Turn Dialogue. DH-RAG is inspired by human cognitive processes that utilize both long-term memory and immediate historical context in conversational responses \citep{stafford1987conversational}. DH-RAG is structured around two principal components: a History-Learning based Query Reconstruction Module, designed to generate effective queries by synthesizing current and prior interactions, and a Dynamic History Information Updating Module, which continually refreshes historical context throughout the dialogue. The center of DH-RAG is a Dynamic Historical Information database, which is further refined by three strategies within the Query Reconstruction Module: Historical Query Clustering, Hierarchical Matching, and Chain of Thought Tracking. Experimental evaluations show that DH-RAG significantly surpasses conventional models on several benchmarks, enhancing response relevance, coherence, and dialogue quality.
Authors: Justin Singh Kang, Landon Butler, Abhineet Agarwal, Yigit Efe Erginbas, Ramtin Pedarsani, Kannan Ramchandran, Bin Yu
Abstract: Large language models (LLMs) have revolutionized machine learning due to their ability to capture complex interactions between input features. Popular post-hoc explanation methods like SHAP provide marginal feature attributions, while their extensions to interaction importances only scale to small input lengths ($\approx 20$). We propose Spectral Explainer (SPEX), a model-agnostic interaction attribution algorithm that efficiently scales to large input lengths ($\approx 1000)$. SPEX exploits underlying natural sparsity among interactions -- common in real-world data -- and applies a sparse Fourier transform using a channel decoding algorithm to efficiently identify important interactions. We perform experiments across three difficult long-context datasets that require LLMs to utilize interactions between inputs to complete the task. For large inputs, SPEX outperforms marginal attribution methods by up to 20% in terms of faithfully reconstructing LLM outputs. Further, SPEX successfully identifies key features and interactions that strongly influence model output. For one of our datasets, HotpotQA, SPEX provides interactions that align with human annotations. Finally, we use our model-agnostic approach to generate explanations to demonstrate abstract reasoning in closed-source LLMs (GPT-4o mini) and compositional reasoning in vision-language models.
Authors: Hui Wang, Zhengpeng Zhao, Jing Wang, Yushu Du, Yuan Cheng, Bing Guo, He Xiao, Chenhao Ma, Xiaomeng Han, Dean You, Jiapeng Guan, Ran Wei, Dawei Yang, Zhe Jiang
Abstract: Deep Neural Networks are increasingly leveraging sparsity to reduce the scaling up of model parameter size. However, reducing wall-clock time through sparsity and pruning remains challenging due to irregular memory access patterns, leading to frequent cache misses. In this paper, we present NPU Vector Runahead (NVR), a prefetching mechanism tailored for NPUs to address cache miss problems in sparse DNN workloads. Rather than optimising memory patterns with high overhead and poor portability, NVR adapts runahead execution to the unique architecture of NPUs. NVR provides a general micro-architectural solution for sparse DNN workloads without requiring compiler or algorithmic support, operating as a decoupled, speculative, lightweight hardware sub-thread alongside the NPU, with minimal hardware overhead (under 5%). NVR achieves an average 90% reduction in cache misses compared to SOTA prefetching in general-purpose processors, delivering 4x average speedup on sparse workloads versus NPUs without prefetching. Moreover, we investigate the advantages of incorporating a small cache (16KB) into the NPU combined with NVR. Our evaluation shows that expanding this modest cache delivers 5x higher performance benefits than increasing the L2 cache size by the same amount.
Authors: Huu-Thien Tran, Phuoc-Sang Pham, Thai-Son Tran, Khoa Luu
Abstract: Referring Multi-Object Tracking (RMOT) is a relatively new concept that has rapidly gained traction as a promising research direction at the intersection of computer vision and natural language processing. Unlike traditional multi-object tracking, RMOT identifies and tracks objects and incorporates textual descriptions for object class names, making the approach more intuitive. Various techniques have been proposed to address this challenging problem; however, most require the training of the entire network due to their end-to-end nature. Among these methods, iKUN has emerged as a particularly promising solution. Therefore, we further explore its pipeline and enhance its performance. In this paper, we introduce a practical module dubbed Memory-Efficient Cross-modality -- MEX. This memory-efficient technique can be directly applied to off-the-shelf trackers like iKUN, resulting in significant architectural improvements. Our method proves effective during inference on a single GPU with 4 GB of memory. Among the various benchmarks, the Refer-KITTI dataset, which offers diverse autonomous driving scenes with relevant language expressions, is particularly useful for studying this problem. Empirically, our method demonstrates effectiveness and efficiency regarding HOTA tracking scores, substantially improving memory allocation and processing speed.
Authors: Jie Zou, Mohammad Aliannejadi, Evangelos Kanoulas, Shuxi Han, Heli Ma, Zheng Wang, Yang Yang, Heng Tao Shen
Abstract: Conversational Product Search (CPS) is confined to simulated conversations due to the lack of real-world CPS datasets that reflect human-like language. Additionally, current conversational datasets are limited to support cross-market and multi-lingual usage. In this paper, we introduce a new CPS data collection protocol and present PSCon, a novel CPS dataset designed to assist product search via human-like conversations. The dataset is constructed using a coached human-to-human data collection protocol and supports two languages and dual markets. Also, the dataset enables thorough exploration of six subtasks of CPS: user intent detection, keyword extraction, system action prediction, question selection, item ranking, and response generation. Furthermore, we also offer an analysis of the dataset and propose a benchmark model on the proposed CPS dataset.
Authors: Dan Zhang, Sining Zhoubian, Min Cai, Fengzu Li, Lekang Yang, Wei Wang, Tianjiao Dong, Ziniu Hu, Jie Tang, Yisong Yue
Abstract: This paper presents DataSciBench, a comprehensive benchmark for evaluating Large Language Model (LLM) capabilities in data science. Recent related benchmarks have primarily focused on single tasks, easily obtainable ground truth, and straightforward evaluation metrics, which limits the scope of tasks that can be evaluated. In contrast, DataSciBench is constructed based on a more comprehensive and curated collection of natural and challenging prompts for uncertain ground truth and evaluation metrics. We develop a semi-automated pipeline for generating ground truth (GT) and validating evaluation metrics. This pipeline utilizes and implements an LLM-based self-consistency and human verification strategy to produce accurate GT by leveraging collected prompts, predefined task types, and aggregate functions (metrics). Furthermore, we propose an innovative Task - Function - Code (TFC) framework to assess each code execution outcome based on precisely defined metrics and programmatic rules. Our experimental framework involves testing 6 API-based models, 8 open-source general models, and 9 open-source code generation models using the diverse set of prompts we have gathered. This approach aims to provide a more comprehensive and rigorous evaluation of LLMs in data science, revealing their strengths and weaknesses. Experimental results demonstrate that API-based models outperform open-sourced models on all metrics and Deepseek-Coder-33B-Instruct achieves the highest score among open-sourced models. We release all code and data at https://github.com/THUDM/DataSciBench.
Authors: Saksham Kiroriwal, Julius Pfrommer, J\"urgen Beyerer
Abstract: To reduce the curse of dimensionality for Gaussian processes (GP), they can be decomposed into a Gaussian Process Network (GPN) of coupled subprocesses with lower dimensionality. In some cases, intermediate observations are available within the GPN. However, intermediate observations are often indirect, noisy, and incomplete in most real-world systems. This work introduces the Partially Observable Gaussian Process Network (POGPN) to model real-world process networks. We model a joint distribution of latent functions of subprocesses and make inferences using observations from all subprocesses. POGPN incorporates observation lenses (observation likelihoods) into the well-established inference method of deep Gaussian processes. We also introduce two training methods for POPGN to make inferences on the whole network using node observations. The application to benchmark problems demonstrates how incorporating partial observations during training and inference can improve the predictive performance of the overall network, offering a promising outlook for its practical application.
Authors: Sein Kim, Hongseok Kang, Kibum Kim, Jiwan Kim, Donghyun Kim, Minchul Yang, Kwangjin Oh, Julian McAuley, Chanyoung Park
Abstract: Large Language Models (LLMs) have recently emerged as promising tools for recommendation thanks to their advanced textual understanding ability and context-awareness. Despite the current practice of training and evaluating LLM-based recommendation (LLM4Rec) models under a sequential recommendation scenario, we found that whether these models understand the sequential information inherent in users' item interaction sequences has been largely overlooked. In this paper, we first demonstrate through a series of experiments that existing LLM4Rec models do not fully capture sequential information both during training and inference. Then, we propose a simple yet effective LLM-based sequential recommender, called LLM-SRec, a method that enhances the integration of sequential information into LLMs by distilling the user representations extracted from a pre-trained CF-SRec model into LLMs. Our extensive experiments show that LLM-SRec enhances LLMs' ability to understand users' item interaction sequences, ultimately leading to improved recommendation performance. Furthermore, unlike existing LLM4Rec models that require fine-tuning of LLMs, LLM-SRec achieves state-of-the-art performance by training only a few lightweight MLPs, highlighting its practicality in real-world applications. Our code is available at https://github.com/Sein-Kim/LLM-SRec.
Authors: Tianyu Guo, Hanlin Zhu, Ruiqi Zhang, Jiantao Jiao, Song Mei, Michael I. Jordan, Stuart Russell
Abstract: "Socrates is human. All humans are mortal. Therefore, Socrates is mortal." This classical example demonstrates two-hop reasoning, where a conclusion logically follows from two connected premises. While transformer-based Large Language Models (LLMs) can make two-hop reasoning, they tend to collapse to random guessing when faced with distracting premises. To understand the underlying mechanism, we train a three-layer transformer on synthetic two-hop reasoning tasks. The training dynamics show two stages: a slow learning phase, where the 3-layer transformer performs random guessing like LLMs, followed by an abrupt phase transitions, where the 3-layer transformer suddenly reaches $100%$ accuracy. Through reverse engineering, we explain the inner mechanisms for how models learn to randomly guess between distractions initially, and how they learn to ignore distractions eventually. We further propose a three-parameter model that supports the causal claims for the mechanisms to the training dynamics of the transformer. Finally, experiments on LLMs suggest that the discovered mechanisms generalize across scales. Our methodologies provide new perspectives for scientific understandings of LLMs and our findings provide new insights into how reasoning emerges during training.
Authors: Shengguang Wu, Fan-Yun Sun, Kaiyue Wen, Nick Haber
Abstract: Recent studies have shown that Large Vision-Language Models (VLMs) tend to neglect image content and over-rely on language-model priors, resulting in errors in visually grounded tasks and hallucinations. We hypothesize that this issue arises because existing VLMs are not explicitly trained to generate texts that are accurately grounded in fine-grained image details. To enhance visual feedback during VLM training, we propose S-VCO (Symmetrical Visual Contrastive Optimization), a novel finetuning objective that steers the model toward capturing important visual details and aligning them with corresponding text tokens. To further facilitate this detailed alignment, we introduce MVC, a paired image-text dataset built by automatically filtering and augmenting visual counterfactual data to challenge the model with hard contrastive cases involving Minimal Visual Contrasts. Experiments show that our method consistently improves VLM performance across diverse benchmarks covering various abilities and domains, achieving up to a 22% reduction in hallucinations, and significant gains in vision-centric and general tasks. Notably, these improvements become increasingly pronounced in benchmarks with higher visual dependency. In short, S-VCO offers a significant enhancement of VLM's visually-dependent task performance while retaining or even improving the model's general abilities. We opensource our code at https://s-vco.github.io/
URLs: https://s-vco.github.io/
Authors: Zeki Doruk Erden, Boi Faltings
Abstract: Current machine learning paradigm relies on continuous representations like neural networks, which iteratively adjust parameters to approximate outcomes rather than directly learning the structure of problem. This spreads information across the network, causing issues like information loss and incomprehensibility Building on prior work in environment dynamics modeling, we propose a method that learns visual space in a structured, continual manner. Our approach refines networks to capture the core structure of objects while representing significant subvariants in structure efficiently. We demonstrate this with 2D shape detection, showing incremental learning on MNIST without overwriting knowledge and creating compact, comprehensible representations. These results offer a promising step toward a transparent, continually learning alternative to traditional neural networks for visual processing.
Authors: Chak Tou Leong, Qingyu Yin, Jian Wang, Wenjie Li
Abstract: The safety alignment of large language models (LLMs) remains vulnerable, as their initial behavior can be easily jailbroken by even relatively simple attacks. Since infilling a fixed template between the input instruction and initial model output is a common practice for existing LLMs, we hypothesize that this template is a key factor behind their vulnerabilities: LLMs' safety-related decision-making overly relies on the aggregated information from the template region, which largely influences these models' safety behavior. We refer to this issue as template-anchored safety alignment. In this paper, we conduct extensive experiments and verify that template-anchored safety alignment is widespread across various aligned LLMs. Our mechanistic analyses demonstrate how it leads to models' susceptibility when encountering inference-time jailbreak attacks. Furthermore, we show that detaching safety mechanisms from the template region is promising in mitigating vulnerabilities to jailbreak attacks. We encourage future research to develop more robust safety alignment techniques that reduce reliance on the template region.
Authors: Guangzhi Xiong, Qiao Jin, Xiao Wang, Yin Fang, Haolin Liu, Yifan Yang, Fangyuan Chen, Zhixing Song, Dengyu Wang, Minjia Zhang, Zhiyong Lu, Aidong Zhang
Abstract: Retrieval-augmented generation (RAG) has shown great potential for knowledge-intensive tasks, but its traditional architectures rely on static retrieval, limiting their effectiveness for complex questions that require sequential information-seeking. While agentic reasoning and search offer a more adaptive approach, most existing methods depend heavily on prompt engineering. In this work, we introduce RAG-Gym, a unified optimization framework that enhances information-seeking agents through fine-grained process supervision at each search step. We also propose ReSearch, a novel agent architecture that synergizes answer reasoning and search query generation within the RAG-Gym framework. Experiments on four challenging datasets show that RAG-Gym improves performance by up to 25.6\% across various agent architectures, with ReSearch consistently outperforming existing baselines. Further analysis highlights the effectiveness of advanced LLMs as process reward judges and the transferability of trained reward models as verifiers for different LLMs. Additionally, we examine the scaling properties of training and inference in agentic RAG. The project homepage is available at https://rag-gym.github.io/.
Authors: Arjun Gupta, Rishik Sathua, Saurabh Gupta
Abstract: Many everyday mobile manipulation tasks require precise interaction with small objects, such as grasping a knob to open a cabinet or pressing a light switch. In this paper, we develop Servoing with Vision Models (SVM), a closed-loop training-free framework that enables a mobile manipulator to tackle such precise tasks involving the manipulation of small objects. SVM employs an RGB-D wrist camera and uses visual servoing for control. Our novelty lies in the use of state-of-the-art vision models to reliably compute 3D targets from the wrist image for diverse tasks and under occlusion due to the end-effector. To mitigate occlusion artifacts, we employ vision models to out-paint the end-effector thereby significantly enhancing target localization. We demonstrate that aided by out-painting methods, open-vocabulary object detectors can serve as a drop-in module to identify semantic targets (e.g. knobs) and point tracking methods can reliably track interaction sites indicated by user clicks. This training-free method obtains an 85% zero-shot success rate on manipulating unseen objects in novel environments in the real world, outperforming an open-loop control method and an imitation learning baseline trained on 1000+ demonstrations by an absolute success rate of 50%.
Authors: Michael Luo, Xiaoxiang Shi, Colin Cai, Tianjun Zhang, Justin Wong, Yichuan Wang, Chi Wang, Yanping Huang, Zhifeng Chen, Joseph E. Gonzalez, Ion Stoica
Abstract: Large language model (LLM) applications are evolving beyond simple chatbots into dynamic, general-purpose agentic programs, which scale LLM calls and output tokens to help AI agents reason, explore, and solve complex tasks. However, existing LLM serving systems ignore dependencies between programs and calls, missing significant opportunities for optimization. Our analysis reveals that programs submitted to LLM serving engines experience long cumulative wait times, primarily due to head-of-line blocking at both the individual LLM request and the program. To address this, we introduce Autellix, an LLM serving system that treats programs as first-class citizens to minimize their end-to-end latencies. Autellix intercepts LLM calls submitted by programs, enriching schedulers with program-level context. We propose two scheduling algorithms-for single-threaded and distributed programs-that preempt and prioritize LLM calls based on their programs' previously completed calls. Our evaluation demonstrates that across diverse LLMs and agentic workloads, Autellix improves throughput of programs by 4-15x at the same latency compared to state-of-the-art systems, such as vLLM.
Authors: Yuan Sui, Yufei He, Nian Liu, Xiaoxin He, Kun Wang, Bryan Hooi
Abstract: Large language models (LLMs) are often challenged by generating erroneous or hallucinated responses, especially in complex reasoning tasks. Leveraging knowledge graphs (KGs) as external knowledge sources has emerged as a viable solution. However, existing KG-enhanced methods, either retrieval-based or agent-based, encounter difficulties in accurately retrieving knowledge and efficiently traversing KGs at scale. In this paper, we propose a unified framework, FiDeLiS, designed to improve the factuality of LLM responses by anchoring answers to verifiable reasoning steps retrieved from a KG. To achieve this, we leverage step-wise beam search with a deductive scoring function, allowing the LLM to validate each reasoning step and halt the search once the question is deducible. In addition, our Path-rag module pre-selects a smaller candidate set for each beam search step, reducing computational costs by narrowing the search space. Extensive experiments show that our training-free and efficient approach outperforms strong baselines, enhancing both factuality and interpretability.
Authors: Chenyan Zhao, Mariana Silva, Seth Poulsen
Abstract: In mathematical proof education, there remains a need for interventions that help students learn to write mathematical proofs. Research has shown that timely feedback can be very helpful to students learning new skills. While for many years natural language processing models have struggled to perform well on tasks related to mathematical texts, recent developments in natural language processing have created the opportunity to complete the task of giving students instant feedback on their mathematical proofs. In this paper, we present a set of training methods and models capable of autograding freeform mathematical proofs by leveraging existing large language models and other machine learning techniques. The models are trained using proof data collected from four different proof by induction problems. We use four different robust large language models to compare their performances, and all achieve satisfactory performances to various degrees. Additionally, we recruit human graders to grade the same proofs as the training data, and find that the best grading model is also more accurate than most human graders. With the development of these grading models, we create and deploy an autograder for proof by induction problems and perform a user study with students. Results from the study shows that students are able to make significant improvements to their proofs using the feedback from the autograder, but students still do not trust the AI autograders as much as they trust human graders. Future work can improve on the autograder feedback and figure out ways to help students trust AI autograders.
Authors: Jun-Yan He, Zhi-Qi Cheng, Chenyang Li, Jingdong Sun, Qi He, Wangmeng Xiang, Hanyuan Chen, Jin-Peng Lan, Xianhui Lin, Kang Zhu, Bin Luo, Yifeng Geng, Xuansong Xie, Alexander G. Hauptmann
Abstract: MetaDesigner introduces a transformative framework for artistic typography synthesis, powered by Large Language Models (LLMs) and grounded in a user-centric design paradigm. Its foundation is a multi-agent system comprising the Pipeline, Glyph, and Texture agents, which collectively orchestrate the creation of customizable WordArt, ranging from semantic enhancements to intricate textural elements. A central feedback mechanism leverages insights from both multimodal models and user evaluations, enabling iterative refinement of design parameters. Through this iterative process, MetaDesigner dynamically adjusts hyperparameters to align with user-defined stylistic and thematic preferences, consistently delivering WordArt that excels in visual quality and contextual resonance. Empirical evaluations underscore the system's versatility and effectiveness across diverse WordArt applications, yielding outputs that are both aesthetically compelling and context-sensitive.
Authors: Haoyang Su, Renqi Chen, Shixiang Tang, Zhenfei Yin, Xinzhe Zheng, Jinzhe Li, Biqing Qi, Qi Wu, Hui Li, Wanli Ouyang, Philip Torr, Bowen Zhou, Nanqing Dong
Abstract: The rapid advancement of scientific progress requires innovative tools that can accelerate knowledge discovery. Although recent AI methods, particularly large language models (LLMs), have shown promise in tasks such as hypothesis generation and experimental design, they fall short of replicating the collaborative nature of real-world scientific practices, where diverse experts work together in teams to tackle complex problems. To address the limitations, we propose an LLM-based multi-agent system, i.e., Virtual Scientists (VirSci), designed to mimic the teamwork inherent in scientific research. VirSci organizes a team of agents to collaboratively generate, evaluate, and refine research ideas. Through comprehensive experiments, we demonstrate that this multi-agent approach outperforms the state-of-the-art method in producing novel scientific ideas. We further investigate the collaboration mechanisms that contribute to its tendency to produce ideas with higher novelty, offering valuable insights to guide future research and illuminating pathways toward building a robust system for autonomous scientific discovery. The code is available at https://github.com/open-sciencelab/Virtual-Scientists.
URLs: https://github.com/open-sciencelab/Virtual-Scientists.
Authors: Kyungmin Min, Minbeom Kim, Kang-il Lee, Dongryeol Lee, Kyomin Jung
Abstract: Large Vision-Language Models (LVLMs) demonstrate impressive capabilities in generating detailed and coherent responses from visual inputs. However, they are prone to generate hallucinations due to an over-reliance on language priors. To address this issue, we investigate the language priors in LVLMs and make two key observations: (1) Even when predicting the tokens associated with image-related part-of-speech (POS), models increasingly rely on linguistic priors as the token sequences grow, thereby amplifying hallucinations. (2) Methods that directly calibrate LVLM's output distribution to mitigate language priors can lead to a degradation in text quality or even exacerbate hallucinations. Based on these findings, we propose a novel method, Summary-Guided Decoding (SumGD). This method naturally encourages the model to focus more on image information by reducing the text context through summaries, while controlling only the image-related POS tokens to maintain text quality. Through experiments, we demonstrate that SumGD achieves state-of-the-art performance on object hallucination benchmarks. Furthermore, in terms of the trade-off between precision and recall, SumGD achieves Pareto optimality among the existing methods. Lastly, we observe that although existing methods struggle to balance the reduction of object hallucinations with maintaining text quality, SumGD demonstrates robustness in handling this challenge.
Authors: Shaina Raza, Caesar Saleh, Emrul Hasan, Franklin Ogidi, Maximus Powers, Veronica Chatrath, Marcelo Lotif, Roya Javadi, Anam Zahid, Vahid Reza Khazaie
Abstract: The integration of Large Language Models (LLMs) and Vision-Language Models (VLMs) opens new avenues for addressing complex challenges in multimodal content analysis, particularly in biased news detection. This study introduces VLBias, a framework that leverages state-of-the-art LLMs and VLMs to detect linguistic and visual biases in news content. We present a multimodal dataset comprising textual content and corresponding images from diverse news sources. We propose a hybrid annotation framework that combines LLM-based annotations with human review to ensure high-quality labeling while reducing costs and enhancing scalability. Our evaluation compares the performance of state-of-the-art SLMs and LLMs for both modalities (text and images) and the results reveal that while SLMs are computationally efficient, LLMs demonstrate superior accuracy in identifying subtle framing and text-visual inconsistencies. Furthermore, empirical analysis shows that incorporating visual cues alongside textual data improves bias detection accuracy by 3 to 5%. This study provides a comprehensive exploration of LLMs, SLMs, and VLMs as tools for detecting multimodal biases in news content and highlights their respective strengths, limitations, and potential for future applications
Authors: Shayne Longpre, Nikhil Singh, Manuel Cherep, Kushagra Tiwary, Joanna Materzynska, William Brannon, Robert Mahari, Naana Obeng-Marnu, Manan Dey, Mohammed Hamdy, Nayan Saxena, Ahmad Mustafa Anis, Emad A. Alghamdi, Vu Minh Chien, Da Yin, Kun Qian, Yizhi Li, Minnie Liang, An Dinh, Shrestha Mohanty, Deividas Mataciunas, Tobin South, Jianguo Zhang, Ariel N. Lee, Campbell S. Lund, Christopher Klamm, Damien Sileo, Diganta Misra, Enrico Shippole, Kevin Klyman, Lester JV Miranda, Niklas Muennighoff, Seonghyeon Ye, Seungone Kim, Vipul Gupta, Vivek Sharma, Xuhui Zhou, Caiming Xiong, Luis Villa, Stella Biderman, Alex Pentland, Sara Hooker, Jad Kabbara
Abstract: Progress in AI is driven largely by the scale and quality of training data. Despite this, there is a deficit of empirical analysis examining the attributes of well-established datasets beyond text. In this work we conduct the largest and first-of-its-kind longitudinal audit across modalities--popular text, speech, and video datasets--from their detailed sourcing trends and use restrictions to their geographical and linguistic representation. Our manual analysis covers nearly 4000 public datasets between 1990-2024, spanning 608 languages, 798 sources, 659 organizations, and 67 countries. We find that multimodal machine learning applications have overwhelmingly turned to web-crawled, synthetic, and social media platforms, such as YouTube, for their training sets, eclipsing all other sources since 2019. Secondly, tracing the chain of dataset derivations we find that while less than 33% of datasets are restrictively licensed, over 80% of the source content in widely-used text, speech, and video datasets, carry non-commercial restrictions. Finally, counter to the rising number of languages and geographies represented in public AI training datasets, our audit demonstrates measures of relative geographical and multilingual representation have failed to significantly improve their coverage since 2013. We believe the breadth of our audit enables us to empirically examine trends in data sourcing, restrictions, and Western-centricity at an ecosystem-level, and that visibility into these questions are essential to progress in responsible AI. As a contribution to ongoing improvements in dataset transparency and responsible use, we release our entire multimodal audit, allowing practitioners to trace data provenance across text, speech, and video.
Authors: Simeon Campos, Henry Papadatos, Fabien Roger, Chlo\'e Touzet, Otter Quarks, Malcolm Murray
Abstract: The recent development of powerful AI systems has highlighted the need for robust risk management frameworks in the AI industry. Although companies have begun to implement safety frameworks, current approaches often lack the systematic rigor found in other high-risk industries. This paper presents a comprehensive risk management framework for the development of frontier AI that bridges this gap by integrating established risk management principles with emerging AI-specific practices. The framework consists of four key components: (1) risk identification (through literature review, open-ended red-teaming, and risk modeling), (2) risk analysis and evaluation using quantitative metrics and clearly defined thresholds, (3) risk treatment through mitigation measures such as containment, deployment controls, and assurance processes, and (4) risk governance establishing clear organizational structures and accountability. Drawing from best practices in mature industries such as aviation or nuclear power, while accounting for AI's unique challenges, this framework provides AI developers with actionable guidelines for implementing robust risk management. The paper details how each component should be implemented throughout the life-cycle of the AI system - from planning through deployment - and emphasizes the importance and feasibility of conducting risk management work prior to the final training run to minimize the burden associated with it.
Authors: Ibrahim Alabdulmohsin, Xiaohua Zhai
Abstract: Recent research in language modeling reveals two scaling effects: the well-known improvement from increased training compute, and a lesser-known boost from applying more sophisticated or computationally intensive inference methods. Inspired by recent findings on the fractal geometry of language, we introduce Recursive INference Scaling (RINS) as a complementary, plug-in recipe for scaling inference time. For a given fixed model architecture and training compute budget, RINS substantially improves language modeling performance. It also generalizes beyond pure language tasks, delivering gains in multimodal systems, including a +2% improvement in 0-shot ImageNet accuracy for SigLIP-B/16. Additionally, by deriving data scaling laws, we show that RINS improves both the asymptotic performance limits and the scaling exponents. These advantages are maintained even when compared to state-of-the-art recursive techniques like the "repeat-all-over" (RAO) strategy in Mobile LLM. Finally, stochastic RINS not only can enhance performance further but also provides the flexibility to optionally forgo increased inference computation at test time with minimal performance degradation.
Authors: Srijith Rajamohan, Ahmed Salhin, Josh Frazier, Rohit Kumar, Yu-Cheng Tsai, Todd Cook
Abstract: The output of Large Language Models (LLMs) are a function of the internal model's parameters and the input provided into the context window. The hypothesis presented here is that under a greedy sampling strategy the variance in the LLM's output is a function of the conceptual certainty embedded in the model's parametric knowledge, as well as the lexical variance in the input. Finetuning the model results in reducing the sensitivity of the model output to the lexical input variations. This is then applied to a classification problem and a probabilistic method is proposed for estimating the certainties of the predicted classes.
Authors: Emre Can Acikgoz, Jeremiah Greer, Akul Datta, Ze Yang, William Zeng, Oussama Elachqar, Emmanouil Koukoumidis, Dilek Hakkani-T\"ur, Gokhan Tur
Abstract: Large Language Models (LLMs) with API-calling capabilities enabled building effective Language Agents (LA), while also revolutionizing the conventional task-oriented dialogue (TOD) paradigm. However, current approaches face a critical dilemma: TOD systems are often trained on a limited set of target APIs, requiring new data to maintain their quality when interfacing with new services, while LAs are not trained to maintain user intent over multi-turn conversations. Because both robust multi-turn management and advanced function calling are crucial for effective conversational agents, we evaluate these skills on three popular benchmarks: MultiWOZ 2.4 (TOD), BFCL V3 (LA), and API-Bank (LA), and our analyses reveal that specialized approaches excel in one domain but underperform in the other. To bridge this chasm, we introduce CoALM (Conversational Agentic Language Model), a unified approach that integrates both conversational and agentic capabilities. We created CoALM-IT, a carefully constructed multi-task dataset that interleave multi-turn ReAct reasoning with complex API usage. Using CoALM-IT, we train three models CoALM 8B, CoALM 70B, and CoALM 405B, which outperform top domain-specific models, including GPT-4o, across all three benchmarks. This demonstrates the feasibility of a single model approach for both TOD and LA, setting a new standard for conversational agents.
Authors: Xinxin You, Xien Liu, Qixin Sun, Huan Zhang, Kaiyin Zhou, Shaohui Liu, GuoPing Hu, ShiJin Wang, Si Liu, Ji Wu
Abstract: Recent methodologies utilizing synthetic datasets have aimed to address inconsistent hallucinations in large language models (LLMs); however,these approaches are primarily tailored to specific tasks, limiting their generalizability. Inspired by the strong performance of code-trained models in logic-intensive domains, we propose a novel framework that leverages event-based text to generate corresponding code and employs cyclic training to transfer the logical consistency of code to natural language effectively. Our method significantly reduces inconsistent hallucinations across three leading LLMs and two categories of natural language tasks while maintaining overall performance. This framework effectively alleviates hallucinations without necessitating adaptation to downstream tasks, demonstrating generality and providing new perspectives to tackle the challenge of inconsistent hallucinations.
Authors: Vardaan Pahuja, Yadong Lu, Corby Rosset, Boyu Gou, Arindam Mitra, Spencer Whitehead, Yu Su, Ahmed Awadallah
Abstract: Recent success in large multimodal models (LMMs) has sparked promising applications of agents capable of autonomously completing complex web tasks. While open-source LMM agents have made significant advances in offline evaluation benchmarks, their performance still falls substantially short of human-level capabilities in more realistic online settings. A key bottleneck is the lack of diverse and large-scale trajectory-level datasets across various domains, which are expensive to collect. In this paper, we address this challenge by developing a scalable recipe to synthesize the largest and most diverse trajectory-level dataset to date, containing over 94K successful multimodal web trajectories, spanning 49K unique URLs, 720K screenshots, and 33M web elements. In particular, we leverage extensive web exploration and refinement to obtain diverse task intents. The average cost is 28 cents per successful trajectory, making it affordable to a wide range of users in the community. Leveraging this dataset, we train Explorer, a multimodal web agent, and demonstrate strong performance on both offline and online web agent benchmarks such as Mind2Web-Live, Multimodal-Mind2Web, and MiniWob++. Additionally, our experiments highlight data scaling as a key driver for improving web agent capabilities. We hope this study makes state-of-the-art LMM-based agent research at a larger scale more accessible.
Authors: Guojun Xiong, Zhiyang Deng, Keyi Wang, Yupeng Cao, Haohang Li, Yangyang Yu, Xueqing Peng, Mingquan Lin, Kaleb E Smith, Xiao-Yang Liu, Jimin Huang, Sophia Ananiadou, Qianqian Xie
Abstract: Large language models (LLMs) fine-tuned on multimodal financial data have demonstrated impressive reasoning capabilities in various financial tasks. However, they often struggle with multi-step, goal-oriented scenarios in interactive financial markets, such as trading, where complex agentic approaches are required to improve decision-making. To address this, we propose \textsc{FLAG-Trader}, a unified architecture integrating linguistic processing (via LLMs) with gradient-driven reinforcement learning (RL) policy optimization, in which a partially fine-tuned LLM acts as the policy network, leveraging pre-trained knowledge while adapting to the financial domain through parameter-efficient fine-tuning. Through policy gradient optimization driven by trading rewards, our framework not only enhances LLM performance in trading but also improves results on other financial-domain tasks. We present extensive empirical evidence to validate these enhancements.
Authors: Shreya Shukla, Jose Torres, Abhijit Mishra, Jacek Gwizdka, Shounak Roychowdhury
Abstract: Integration of Brain-Computer Interfaces (BCIs) and Generative Artificial Intelligence (GenAI) has opened new frontiers in brain signal decoding, enabling assistive communication, neural representation learning, and multimodal integration. BCIs, particularly those leveraging Electroencephalography (EEG), provide a non-invasive means of translating neural activity into meaningful outputs. Recent advances in deep learning, including Generative Adversarial Networks (GANs) and Transformer-based Large Language Models (LLMs), have significantly improved EEG-based generation of images, text, and speech. This paper provides a literature review of the state-of-the-art in EEG-based multimodal generation, focusing on (i) EEG-to-image generation through GANs, Variational Autoencoders (VAEs), and Diffusion Models, and (ii) EEG-to-text generation leveraging Transformer based language models and contrastive learning methods. Additionally, we discuss the emerging domain of EEG-to-speech synthesis, an evolving multimodal frontier. We highlight key datasets, use cases, challenges, and EEG feature encoding methods that underpin generative approaches. By providing a structured overview of EEG-based generative AI, this survey aims to equip researchers and practitioners with insights to advance neural decoding, enhance assistive technologies, and expand the frontiers of brain-computer interaction.
Authors: Hyejun Jeong, Hamin Son, Seohu Lee, Jayun Hyun, Tai-Myoung Chung
Abstract: Federated learning is a distributed framework designed to address privacy concerns. However, it introduces new attack surfaces, which are especially prone when data is non-Independently and Identically Distributed. Existing approaches fail to effectively mitigate the malicious influence in this setting; previous approaches often tackle non-IID data and poisoning attacks separately. To address both challenges simultaneously, we present FedCC, a simple yet effective novel defense algorithm against model poisoning attacks. It leverages the Centered Kernel Alignment similarity of Penultimate Layer Representations for clustering, allowing the identification and filtration of malicious clients, even in non-IID data settings. The penultimate layer representations are meaningful since the later layers are more sensitive to local data distributions, which allows better detection of malicious clients. The sophisticated utilization of layer-wise Centered Kernel Alignment similarity allows attack mitigation while leveraging useful knowledge obtained. Our extensive experiments demonstrate the effectiveness of FedCC in mitigating both untargeted model poisoning and targeted backdoor attacks. Compared to existing outlier detection-based and first-order statistics-based methods, FedCC consistently reduces attack confidence to zero. Specifically, it significantly minimizes the average degradation of global performance by 65.5\%. We believe that this new perspective on aggregation makes it a valuable contribution to the field of FL model security and privacy. The code will be made available upon acceptance.
Authors: Noah A. Crum, Leanto Sunny, Pooya Ronagh, Raymond Laflamme, Radhakrishnan Balu, George Siopsis
Abstract: Motivated by applications of quantum computers in Gibbs sampling from continuous real-valued functions, we ask whether such algorithms can provide practical advantages for machine learning models trained on classical data and seek measures for quantifying such impacts. In this study, we focus on deep energy-based models (EBM), as they require continuous-domain Gibbs sampling both during training and inference. In lieu of fault-tolerant quantum computers that can execute quantum Gibbs sampling algorithms, we use the Monte Carlo simulation of diffusion processes as a classical alternative. More specifically, we investigate whether long-run persistent chain Monte Carlo simulation of Langevin dynamics improves the quality of the representations achieved by EBMs. We consider a scheme in which the Monte Carlo simulation of a diffusion, whose drift is given by the gradient of the energy function, is used to improve the adversarial robustness and calibration score of an independent classifier network. Our results show that increasing the computational budget of Gibbs sampling in persistent contrastive divergence improves both the calibration and adversarial robustness of the model, suggesting a prospective avenue of quantum advantage for generative AI using future large-scale quantum computers.
Authors: Emanuele Ballarin, Alessio Ansuini, Luca Bortolussi
Abstract: In this work, we propose a novel adversarial defence mechanism for image classification - CARSO - blending the paradigms of adversarial training and adversarial purification in a synergistic robustness-enhancing way. The method builds upon an adversarially-trained classifier, and learns to map its internal representation associated with a potentially perturbed input onto a distribution of tentative clean reconstructions. Multiple samples from such distribution are classified by the same adversarially-trained model, and a carefully chosen aggregation of its outputs finally constitutes the robust prediction of interest. Experimental evaluation by a well-established benchmark of strong adaptive attacks, across different image datasets, shows that CARSO is able to defend itself against adaptive end-to-end white-box attacks devised for stochastic defences. Paying a modest clean accuracy toll, our method improves by a significant margin the state-of-the-art for Cifar-10, Cifar-100, and TinyImageNet-200 $\ell_\infty$ robust classification accuracy against AutoAttack. Code, and instructions to obtain pre-trained models are available at: https://github.com/emaballarin/CARSO .
Authors: Jae Yong Lee, Seungchan Ko, Youngjoon Hong
Abstract: Partial differential equations (PDEs) underlie our understanding and prediction of natural phenomena across numerous fields, including physics, engineering, and finance. However, solving parametric PDEs is a complex task that necessitates efficient numerical methods. In this paper, we propose a novel approach for solving parametric PDEs using a Finite Element Operator Network (FEONet). Our proposed method leverages the power of deep learning in conjunction with traditional numerical methods, specifically the finite element method, to solve parametric PDEs in the absence of any paired input-output training data. We performed various experiments on several benchmark problems and confirmed that our approach has demonstrated excellent performance across various settings and environments, proving its versatility in terms of accuracy, generalization, and computational flexibility. While our method is not meshless, the FEONet framework shows potential for application in various fields where PDEs play a crucial role in modeling complex domains with diverse boundary conditions and singular behavior. Furthermore, we provide theoretical convergence analysis to support our approach, utilizing finite element approximation in numerical analysis.
Authors: Abdellah Rahmani, Pascal Frossard
Abstract: Understanding causal relationships in multivariate time series is essential for predicting and controlling dynamic systems in fields like economics, neuroscience, and climate science. However, existing causal discovery methods often assume stationarity, limiting their effectiveness when time series consist of sequential regimes, consecutive temporal segments with unknown boundaries and changing causal structures. In this work, we firstly introduce a framework to describe and model such time series. Then, we present CASTOR, a novel method that concurrently learns the Directed Acyclic Graph (DAG) for each regime while determining the number of regimes and their sequential arrangement. CASTOR optimizes the data log-likelihood using an expectation-maximization algorithm, alternating between assigning regime indices (expectation step) and inferring causal relationships in each regime (maximization step). We establish the identifiability of the regimes and DAGs within our framework. Extensive experiments show that CASTOR consistently outperforms existing causal discovery models in detecting different regimes and learning their DAGs across various settings, including linear and nonlinear causal relationships, on both synthetic and real world datasets.
Authors: Nan Huang, Ting Zhang, Yuhui Yuan, Dong Chen, Shanghang Zhang
Abstract: In this paper, we address the critical bottleneck in robotics caused by the scarcity of diverse 3D data by presenting a novel two-stage approach for generating high-quality 3D models from a single image. This method is motivated by the need to efficiently expand 3D asset creation, particularly for robotics datasets, where the variety of object types is currently limited compared to general image datasets. Unlike previous methods that primarily rely on general diffusion priors, which often struggle to align with the reference image, our approach leverages subject-specific prior knowledge. By incorporating subject-specific priors in both geometry and texture, we ensure precise alignment between the generated 3D content and the reference object. Specifically, we introduce a shading mode-aware prior into the NeRF optimization process, enhancing the geometry and refining texture in the coarse outputs to achieve superior quality. Extensive experiments demonstrate that our method significantly outperforms prior approaches.
Authors: Karina Corti\~nas-Lorenzo, Si\^an Lindley, Ida Larsen-Ledet, Bhaskar Mitra
Abstract: Knowledge can't be disentangled from people. As AI knowledge systems mine vast volumes of work-related data, the knowledge that's being extracted and surfaced is intrinsically linked to the people who create and use it. When predictive algorithms that learn from data are used to link knowledge and people, inaccuracies in knowledge extraction and surfacing can lead to disproportionate harms, influencing how individuals see each other and how they see themselves at work. In this paper, we present a reflective analysis of transparency requirements and impacts in this type of systems. We conduct a multidisciplinary literature review to understand the impacts of transparency in workplace settings, introducing the looking-glass metaphor to conceptualize AI knowledge systems as systems that reflect and distort, expanding our view on transparency requirements, implications and challenges. We formulate transparency as a key mediator in shaping different ways of seeing, including seeing into the system, which unveils its capabilities, limitations and behavior, and seeing through the system, which shapes workers' perceptions of their own contributions and others within the organization. Recognizing the sociotechnical nature of these systems, we identify three transparency dimensions necessary to realize the value of AI knowledge systems, namely system transparency, procedural transparency and transparency of outcomes. We discuss key challenges hindering the implementation of these forms of transparency, bringing to light the wider sociotechnical gap and highlighting directions for future Computer-supported Cooperative Work (CSCW) research.
Authors: David Hartmann, Jos\'e Renato Laranjeira de Pereira, Chiara Streitb\"orger, Bettina Berendt
Abstract: The European legislature has proposed the Digital Services Act (DSA) and Artificial Intelligence Act (AIA) to regulate platforms and Artificial Intelligence (AI) products. We review to what extent third-party audits are part of both laws and how is access to information on models and the data provided. By considering the value of third-party audits and third-party data access in an audit ecosystem, we identify a regulatory gap in that the AIA does not provide access to data for researchers and civil society. Our contributions to the literature include: (1) Defining an AI audit ecosystem incorporating compliance and oversight. (2) Highlighting a regulatory gap within the DSA and AIA regulatory framework, preventing the establishment of an AI audit ecosystem that has effective oversight by civil society and academia. (3) Emphasizing that third-party audits by research and civil society must be part of that ecosystem, we call for AIA amendments and delegated acts to include data and model access for certain AI products. Furthermore, we call for the DSA to provide NGOs and investigative journalists with data access to platforms by delegated acts and for adaptions and amendments of the AIA to provide third-party audits and data and model access, at least for high-risk systems. Regulations modeled after EU AI regulations should enable data access and third-party audits, fostering an AI audit ecosystem that promotes compliance and oversight mechanisms.
Authors: Jiang Li, Xiangdong Su, Guanglai Gao
Abstract: Recent studies have highlighted the effectiveness of tensor decomposition methods in the Temporal Knowledge Graphs Embedding (TKGE) task. However, we found that inherent heterogeneity among factor tensors in tensor decomposition significantly hinders the tensor fusion process and further limits the performance of link prediction. To overcome this limitation, we introduce a novel method that maps factor tensors onto a unified smooth Lie group manifold to make the distribution of factor tensors approximating homogeneous in tensor decomposition. We provide the theoretical proof of our motivation that homogeneous tensors are more effective than heterogeneous tensors in tensor fusion and approximating the target for tensor decomposition based TKGE methods. The proposed method can be directly integrated into existing tensor decomposition based TKGE methods without introducing extra parameters. Extensive experiments demonstrate the effectiveness of our method in mitigating the heterogeneity and in enhancing the tensor decomposition based TKGE models.
Authors: Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, Dasha Metropolitansky, Robert Osazuwa Ness, Jonathan Larson
Abstract: The use of retrieval-augmented generation (RAG) to retrieve relevant information from an external knowledge source enables large language models (LLMs) to answer questions over private and/or previously unseen document collections. However, RAG fails on global questions directed at an entire text corpus, such as "What are the main themes in the dataset?", since this is inherently a query-focused summarization (QFS) task, rather than an explicit retrieval task. Prior QFS methods, meanwhile, do not scale to the quantities of text indexed by typical RAG systems. To combine the strengths of these contrasting methods, we propose GraphRAG, a graph-based approach to question answering over private text corpora that scales with both the generality of user questions and the quantity of source text. Our approach uses an LLM to build a graph index in two stages: first, to derive an entity knowledge graph from the source documents, then to pregenerate community summaries for all groups of closely related entities. Given a question, each community summary is used to generate a partial response, before all partial responses are again summarized in a final response to the user. For a class of global sensemaking questions over datasets in the 1 million token range, we show that GraphRAG leads to substantial improvements over a conventional RAG baseline for both the comprehensiveness and diversity of generated answers.
Authors: Zhili Liu, Yunhao Gou, Kai Chen, Lanqing Hong, Jiahui Gao, Fei Mi, Yu Zhang, Zhenguo Li, Xin Jiang, Qun Liu, James T. Kwok
Abstract: As the capabilities of large language models (LLMs) continue to expand, aligning these models with human values remains a significant challenge. Recent studies show that reasoning abilities contribute significantly to model safety, while integrating Mixture-of-Experts (MoE) architectures can further enhance alignment. In this work, we propose Mixture of insighTful Experts (MoTE), a novel framework that synergistically combines reasoning chains and expert mixtures to improve self-alignments. From a data perspective, MoTE employs a structured reasoning chain comprising four key stages: Question Analysis, Answer Guidance, Safe Answer, and Safety Checking. This approach enhances safety through multi-step reasoning and proves effective even for smaller and less powerful LLMs (e.g., 7B models). From an architectural perspective, MoTE adopts a multi-LoRA framework with step-level routing, where each expert is dedicated to a specific reasoning step. This design eliminates the need for balance losses, ensures stable training, and supports adaptive inference lengths. Experimental results demonstrate that MoTE significantly improves model safety, jailbreak resistance, and over-refusal capabilities, achieving performance comparable to OpenAI's state-of-the-art o1 model.
Authors: Yizhuo Lu, Changde Du, Chong Wang, Xuanliu Zhu, Liuyun Jiang, Xujin Li, Huiguang He
Abstract: Reconstructing human dynamic vision from brain activity is a challenging task with great scientific significance. Although prior video reconstruction methods have made substantial progress, they still suffer from several limitations, including: (1) difficulty in simultaneously reconciling semantic (e.g. categorical descriptions), structure (e.g. size and color), and consistent motion information (e.g. order of frames); (2) low temporal resolution of fMRI, which poses a challenge in decoding multiple frames of video dynamics from a single fMRI frame; (3) reliance on video generation models, which introduces ambiguity regarding whether the dynamics observed in the reconstructed videos are genuinely derived from fMRI data or are hallucinations from generative model. To overcome these limitations, we propose a two-stage model named Mind-Animator. During the fMRI-to-feature stage, we decouple semantic, structure, and motion features from fMRI. Specifically, we employ fMRI-vision-language tri-modal contrastive learning to decode semantic feature from fMRI and design a sparse causal attention mechanism for decoding multi-frame video motion features through a next-frame-prediction task. In the feature-to-video stage, these features are integrated into videos using an inflated Stable Diffusion, effectively eliminating external video data interference. Extensive experiments on multiple video-fMRI datasets demonstrate that our model achieves state-of-the-art performance. Comprehensive visualization analyses further elucidate the interpretability of our model from a neurobiological perspective. Project page: https://mind-animator-design.github.io/.
Authors: Gabriele Dominici, Pietro Barbiero, Mateo Espinosa Zarlenga, Alberto Termine, Martin Gjoreski, Giuseppe Marra, Marc Langheinrich
Abstract: Causal opacity denotes the difficulty in understanding the "hidden" causal structure underlying the decisions of deep neural network (DNN) models. This leads to the inability to rely on and verify state-of-the-art DNN-based systems, especially in high-stakes scenarios. For this reason, circumventing causal opacity in DNNs represents a key open challenge at the intersection of deep learning, interpretability, and causality. This work addresses this gap by introducing Causal Concept Graph Models (Causal CGMs), a class of interpretable models whose decision-making process is causally transparent by design. Our experiments show that Causal CGMs can: (i) match the generalisation performance of causally opaque models, (ii) enable human-in-the-loop corrections to mispredicted intermediate reasoning steps, boosting not just downstream accuracy after corrections but also the reliability of the explanations provided for specific instances, and (iii) support the analysis of interventional and counterfactual scenarios, thereby improving the model's causal interpretability and supporting the effective verification of its reliability and fairness.
Authors: Shicong Cen, Jincheng Mei, Katayoon Goshvadi, Hanjun Dai, Tong Yang, Sherry Yang, Dale Schuurmans, Yuejie Chi, Bo Dai
Abstract: Reinforcement learning from human feedback (RLHF) has demonstrated great promise in aligning large language models (LLMs) with human preference. Depending on the availability of preference data, both online and offline RLHF are active areas of investigation. A key bottleneck is understanding how to incorporate uncertainty estimation in the reward function learned from the preference data for RLHF, regardless of how the preference data is collected. While the principles of optimism or pessimism under uncertainty are well-established in standard reinforcement learning (RL), a practically-implementable and theoretically-grounded form amenable to large language models is not yet available, as standard techniques for constructing confidence intervals become intractable under arbitrary policy parameterizations. In this paper, we introduce a unified approach to online and offline RLHF -- value-incentivized preference optimization (VPO) -- which regularizes the maximum-likelihood estimate of the reward function with the corresponding value function, modulated by a $\textit{sign}$ to indicate whether the optimism or pessimism is chosen. VPO also directly optimizes the policy with implicit reward modeling, and therefore shares a simpler RLHF pipeline similar to direct preference optimization. Theoretical guarantees of VPO are provided for both online and offline settings, matching the rates of their standard RL counterparts. Moreover, experiments on text summarization and dialog verify the practicality and effectiveness of VPO.
Authors: Ruizhong Qiu, Weiliang Will Zeng, James Ezick, Christopher Lott, Hanghang Tong
Abstract: The emergence of large language models (LLMs) has significantly pushed the frontiers of program synthesis. Advancement of LLM-based program synthesis calls for a thorough evaluation of LLM-generated code. Most evaluation frameworks focus on the (functional) correctness of generated code; efficiency, as an important measure of code quality, has been overlooked in existing evaluations. In this work, we develop ENAMEL (EfficeNcy AutoMatic EvaLuator), a rigorous and high-standard benchmark for evaluating the capability of LLMs in generating efficient code. Firstly, we propose a new efficiency metric called eff@k, which generalizes the pass@k metric from correctness to efficiency and appropriately handles right-censored execution time. Furthermore, we derive an unbiased and variance-reduced estimator of eff@k via Rao--Blackwellization; we also provide a numerically stable implementation for the new estimator. Secondly, to set a high-standard for efficiency evaluation, we employ a human expert to design best algorithms and implementations as our reference solutions of efficiency, many of which are much more efficient than existing canonical solutions in HumanEval and HumanEval+. Moreover, to ensure a rigorous evaluation, we employ a human expert to curate strong test case generators to filter out wrong code and differentiate suboptimal algorithms. An extensive study across 30 popular LLMs using our benchmark ENAMEL shows that LLMs still fall short of generating expert-level efficient code. Using two subsets of our problem set, we demonstrate that such deficiency is because current LLMs struggle in designing advanced algorithms and are barely aware of implementation optimization. Our benchmark is publicly available at https://github.com/q-rz/enamel .
Authors: Jing Gong, Yanghui Wu, Linxi Liang, Jiachi Chen, Mingwei Liu, Yanlin Wang, Zibin Zheng
Abstract: Semantic code search, retrieving code that matches a given natural language query, is an important task to improve productivity in software engineering. Existing code search datasets face limitations: they rely on human annotators who assess code primarily through semantic understanding rather than functional verification, leading to potential inaccuracies and scalability issues. Additionally, current evaluation metrics often overlook the multi-choice nature of code search. This paper introduces CoSQA+, pairing high-quality queries from CoSQA with multiple suitable codes. We develop an automated pipeline featuring multiple model-based candidate selections and the novel test-driven agent annotation system. Among a single Large Language Model (LLM) annotator and Python expert annotators (without test-based verification), agents leverage test-based verification and achieve the highest accuracy of 96.4%. Through extensive experiments, CoSQA+ has demonstrated superior quality over CoSQA. Models trained on CoSQA+ exhibit improved performance. We provide the code and data at https://github.com/DeepSoftwareAnalytics/CoSQA_Plus.
Authors: Yuchen Wen, Keping Bi, Wei Chen, Jiafeng Guo, Xueqi Cheng
Abstract: As large language models (LLMs) become an important way of information access, there have been increasing concerns that LLMs may intensify the spread of unethical content, including implicit bias that hurts certain populations without explicit harmful words. In this paper, we conduct a rigorous evaluation of LLMs' implicit bias towards certain demographics by attacking them from a psychometric perspective to elicit agreements to biased viewpoints. Inspired by psychometric principles in cognitive and social psychology, we propose three attack approaches, i.e., Disguise, Deception, and Teaching. Incorporating the corresponding attack instructions, we built two benchmarks: (1) a bilingual dataset with biased statements covering four bias types (2.7K instances) for extensive comparative analysis, and (2) BUMBLE, a larger benchmark spanning nine common bias types (12.7K instances) for comprehensive evaluation. Extensive evaluation of popular commercial and open-source LLMs shows that our methods can elicit LLMs' inner bias more effectively than competitive baselines. Our attack methodology and benchmarks offer an effective means of assessing the ethical risks of LLMs, driving progress toward greater accountability in their development.
Authors: Hongbo Li, Sen Lin, Lingjie Duan, Yingbin Liang, Ness B. Shroff
Abstract: Continual learning (CL) has garnered significant attention because of its ability to adapt to new tasks that arrive over time. Catastrophic forgetting (of old tasks) has been identified as a major issue in CL, as the model adapts to new tasks. The Mixture-of-Experts (MoE) model has recently been shown to effectively mitigate catastrophic forgetting in CL, by employing a gating network to sparsify and distribute diverse tasks among multiple experts. However, there is a lack of theoretical analysis of MoE and its impact on the learning performance in CL. This paper provides the first theoretical results to characterize the impact of MoE in CL via the lens of overparameterized linear regression tasks. We establish the benefit of MoE over a single expert by proving that the MoE model can diversify its experts to specialize in different tasks, while its router learns to select the right expert for each task and balance the loads across all experts. Our study further suggests an intriguing fact that the MoE in CL needs to terminate the update of the gating network after sufficient training rounds to attain system convergence, which is not needed in the existing MoE studies that do not consider the continual task arrival. Furthermore, we provide explicit expressions for the expected forgetting and overall generalization error to characterize the benefit of MoE in the learning performance in CL. Interestingly, adding more experts requires additional rounds before convergence, which may not enhance the learning performance. Finally, we conduct experiments on both synthetic and real datasets to extend these insights from linear models to deep neural networks (DNNs), which also shed light on the practical algorithm design for MoE in CL.
Authors: Erik B. Terres-Escudero, Javier Del Ser, Aitor Mart\'inez-Seras, Pablo Garcia-Bringas
Abstract: In recent years, Artificial Intelligence (AI) models have achieved remarkable success across various domains, yet challenges persist in two critical areas: ensuring robustness against uncertain inputs and drastically increasing model efficiency during training and inference. Spiking Neural Networks (SNNs), inspired by biological systems, offer a promising avenue for overcoming these limitations. By operating in an event-driven manner, SNNs achieve low energy consumption and can naturally implement biological methods known for their high noise tolerance. In this work, we explore the potential of the spiking Forward-Forward Algorithm (FFA) to address these challenges, leveraging its representational properties for both Out-of-Distribution (OoD) detection and interpretability. To achieve this, we exploit the sparse and highly specialized neural latent space of FF networks to estimate the likelihood of a sample belonging to the training distribution. Additionally, we propose a novel, gradient-free attribution method to detect features that drive a sample away from class distributions, addressing the challenges posed by the lack of gradients in most visual interpretability methods for spiking models. We evaluate our OoD detection algorithm on well-known image datasets (e.g., Omniglot, Not-MNIST, CIFAR10), outperforming previous methods proposed in the recent literature for OoD detection in spiking networks. Furthermore, our attribution method precisely identifies salient OoD features, such as artifacts or missing regions, hence providing a visual explanatory interface for the user to understand why unknown inputs are identified as such by the proposed method.
Authors: Shih-Kai Chou, Jernej Hribar, Vid Han\v{z}el, Mihael Mohor\v{c}i\v{c}, Carolina Fortuna
Abstract: Artificial Intelligence (AI) coupled with the existing Internet of Things (IoT) enables more autonomous operations across various economic sectors. While this paradigm shift results in increased energy consumption it is difficult to quantify the end-to-end energy consumption of such systems with the conventional metrics as they either focus on the communication, the computation infrastructure or model development. To address this, we propose a new metric, the Energy Cost of AI lifecycle (eCAL). eCAL captures the energy consumption throughout the architectural components and lifecycle of an AI-powered wireless system by analyzing the complexity of data collection and manipulation in individual components and deriving overall and per-bit energy consumption. We show that the better a model and the more it is used, the more energy efficient an inference is. For an example Artificial Intelligence of Things (AIoT) configuration, eCAL for making 100 inferences is 2.73 times higher than for 1000 inferences. Additionally, we developed a modular open source simulation tool to enable researchers, practitioners, and engineers to calculate the end-to-end energy cost with various configurations and across various systems, ensuring adaptability to diverse use cases.
Authors: Xiaohua Feng, Yuyuan Li, Chaochao Chen, Li Zhang, Longfei Li, Jun Zhou, Xiaolin Zheng
Abstract: While generative models have made significant advancements in recent years, they also raise concerns such as privacy breaches and biases. Machine unlearning has emerged as a viable solution, aiming to remove specific training data, e.g., containing private information and bias, from models. In this paper, we study the machine unlearning problem in Image-to-Image (I2I) generative models. Previous studies mainly treat it as a single objective optimization problem, offering a solitary solution, thereby neglecting the varied user expectations towards the trade-off between complete unlearning and model utility. To address this issue, we propose a controllable unlearning framework that uses a control coefficient $\varepsilon$ to control the trade-off. We reformulate the I2I generative model unlearning problem into a $\varepsilon$-constrained optimization problem and solve it with a gradient-based method to find optimal solutions for unlearning boundaries. These boundaries define the valid range for the control coefficient. Within this range, every yielded solution is theoretically guaranteed with Pareto optimality. We also analyze the convergence rate of our framework under various control functions. Extensive experiments on two benchmark datasets across three mainstream I2I models demonstrate the effectiveness of our controllable unlearning framework.
Authors: Ritwik Mishra, Sreeram Vennam, Rajiv Ratn Shah, Ponnurangam Kumaraguru
Abstract: Most existing Question Answering Datasets (QuADs) primarily focus on factoid-based short-context Question Answering (QA) in high-resource languages. However, the scope of such datasets for low-resource languages remains limited, with only a few works centered on factoid-based QuADs and none on non-factoid QuADs. Therefore, this work presents MuNfQuAD, a multilingual QuAD with non-factoid questions. It utilizes interrogative sub-headings from BBC news articles as questions and the corresponding paragraphs as silver answers. The dataset comprises over 578K QA pairs across 38 languages, encompassing several low-resource languages, and stands as the largest multilingual QA dataset to date. Based on the manual annotations of 790 QA-pairs from MuNfQuAD (golden set), we observe that 98\% of questions can be answered using their corresponding silver answer. Our fine-tuned Answer Paragraph Selection (APS) model outperforms the baselines. The APS model attained an accuracy of 80\% and 72\%, as well as a macro F1 of 72\% and 66\%, on the MuNfQuAD testset and the golden set, respectively. Furthermore, the APS model effectively generalizes a certain language within the golden set, even after being fine-tuned on silver labels. We also observe that the fine-tuned APS model is beneficial for reducing the context of a question. These findings suggest that this resource would be a valuable contribution to the QA research community.
Authors: Shen Li, Liuyi Yao, Lan Zhang, Yaliang Li
Abstract: Aligned LLMs are secure, capable of recognizing and refusing to answer malicious questions. However, the role of internal parameters in maintaining such security is not well understood yet, further these models can be vulnerable to security degradation when subjected to fine-tuning attacks. To address these challenges, our work uncovers the mechanism behind security in aligned LLMs at the parameter level, identifying a small set of contiguous layers in the middle of the model that are crucial for distinguishing malicious queries from normal ones, referred to as ``safety layers". We first confirm the existence of these safety layers by analyzing variations in input vectors within the model's internal layers. Additionally, we leverage the over-rejection phenomenon and parameters scaling analysis to precisely locate the safety layers. Building on these findings, we propose a novel fine-tuning approach, Safely Partial-Parameter Fine-Tuning (SPPFT), that fixes the gradient of the safety layers during fine-tuning to address the security degradation. Our experiments demonstrate that the proposed approach can significantly preserve LLM security while maintaining performance and reducing computational resources compared to full fine-tuning.
Authors: RuiKang OuYang, Bo Qiang, Zixing Song, Jos\'e Miguel Hern\'andez-Lobato
Abstract: Developing an efficient sampler capable of generating independent and identically distributed (IID) samples from a Boltzmann distribution is a crucial challenge in scientific research, e.g. molecular dynamics. In this work, we intend to learn neural samplers given energy functions instead of data sampled from the Boltzmann distribution. By learning the energies of the noised data, we propose a diffusion-based sampler, Noised Energy Matching, which theoretically has lower variance and more complexity compared to related works. Furthermore, a novel bootstrapping technique is applied to NEM to balance between bias and variance. We evaluate NEM and BNEM on a 2-dimensional 40 Gaussian Mixture Model (GMM) and a 4-particle double-well potential (DW-4). The experimental results demonstrate that BNEM can achieve state-of-the-art performance while being more robust.
Authors: Carlos E. Luis, Alessandro G. Bottero, Julia Vinogradska, Felix Berkenkamp, Jan Peters
Abstract: Optimal decision-making under partial observability requires reasoning about the uncertainty of the environment's hidden state. However, most reinforcement learning architectures handle partial observability with sequence models that have no internal mechanism to incorporate uncertainty in their hidden state representation, such as recurrent neural networks, deterministic state-space models and transformers. Inspired by advances in probabilistic world models for reinforcement learning, we propose a standalone Kalman filter layer that performs closed-form Gaussian inference in linear state-space models and train it end-to-end within a model-free architecture to maximize returns. Similar to efficient linear recurrent layers, the Kalman filter layer processes sequential data using a parallel scan, which scales logarithmically with the sequence length. By design, Kalman filter layers are a drop-in replacement for other recurrent layers in standard model-free architectures, but importantly they include an explicit mechanism for probabilistic filtering of the latent state representation. Experiments in a wide variety of tasks with partial observability show that Kalman filter layers excel in problems where uncertainty reasoning is key for decision-making, outperforming other stateful models.
Authors: Zhuochun Li, Yuelyu Ji, Rui Meng, Daqing He
Abstract: While reasoning capabilities typically emerge in large language models (LLMs) with tens of billions of parameters, recent research focuses on improving smaller open-source models through knowledge distillation (KD) from commercial LLMs. However, many of these studies rely solely on responses from a single LLM as the gold rationale, unlike the natural human learning process, which involves understanding both the correct answers and the reasons behind mistakes. In this paper, we introduce a novel Fault-Aware DistIllation via Peer-Review (FAIR) approach: 1) Instead of merely obtaining rationales from teachers, our method asks teachers to identify and explain the student's mistakes, providing customized instruction learning data. 2) We design a simulated peer-review process between teacher LLMs, which selects only the generated rationales above the acceptance threshold. This reduces the chance of teachers guessing correctly with flawed rationale, improving instructional data quality. Comprehensive experiments and analysis on mathematical, commonsense, and logical reasoning tasks demonstrate the effectiveness of our method.
Authors: Chang Zou, Xuyang Liu, Ting Liu, Siteng Huang, Linfeng Zhang
Abstract: Diffusion transformers have shown significant effectiveness in both image and video synthesis at the expense of huge computation costs. To address this problem, feature caching methods have been introduced to accelerate diffusion transformers by caching the features in previous timesteps and reusing them in the following timesteps. However, previous caching methods ignore that different tokens exhibit different sensitivities to feature caching, and feature caching on some tokens may lead to 10$\times$ more destruction to the overall generation quality compared with other tokens. In this paper, we introduce token-wise feature caching, allowing us to adaptively select the most suitable tokens for caching, and further enable us to apply different caching ratios to neural layers in different types and depths. Extensive experiments on PixArt-$\alpha$, OpenSora, and DiT demonstrate our effectiveness in both image and video generation with no requirements for training. For instance, 2.36$\times$ and 1.93$\times$ acceleration are achieved on OpenSora and PixArt-$\alpha$ with almost no drop in generation quality.
Authors: Mutian He, Philip N. Garner
Abstract: Architectures such as Linformer and Mamba have recently emerged as competitive linear time replacements for transformers. However, corresponding large pretrained models are often unavailable, especially in non-text domains. To remedy this, we present a Cross-Architecture Layerwise Distillation (CALD) approach that jointly converts a transformer model to a linear time substitute and fine-tunes it to a target task. We also compare several means to guide the fine-tuning to optimally retain the desired inference capability from the original model. The methods differ in their use of the target model and the trajectory of the parameters. In a series of empirical studies on language processing, language modeling, and speech processing, we show that CALD can effectively recover the result of the original model, and that the guiding strategy contributes to the result. Some reasons for the variation are suggested.
Authors: Yuan Sui, Yufei He, Zifeng Ding, Bryan Hooi
Abstract: Recent works integrating Knowledge Graphs (KGs) have led to promising improvements in enhancing the reasoning accuracy of Large Language Models (LLMs). However, current benchmarks focus mainly on closed-ended tasks, leaving a gap in the assessment of more complex real-world scenarios. This gap has also obscured the evaluation of KGs' potential to mitigate the problem of hallucination in LLMs. To fill the gap, we introduce OKGQA, a new benchmark specifically designed to assess LLMs enhanced with KGs under open-ended, real-world question answering scenarios. OKGQA is designed to closely reflect the complexities of practical applications using questions from different types, and incorporates specific metrics to measure both hallucination ratio and the enhancement in reasoning capabilities. To consider the scenario in which KGs may have varying levels of mistakes, we propose another benchmark variant OKGQA-P to assess model performance when the semantics and structure of KGs are deliberately perturbed and contaminated. OKGQA aims to (1) explore whether KGs can make LLMs more trustworthy in an open-ended setting, and (2) conduct a comparative analysis to shed light on method design. We believe that this study can facilitate a more complete performance comparison and encourage continuous improvement in integrating KGs with LLMs to reduce hallucination.
Authors: Weinan Zhang, Junwei Liao, Ning Li, Kounianhua Du, Jianghao Lin
Abstract: Since the 1970s, information retrieval (IR) has long been defined as the process of acquiring relevant information items from a pre-defined corpus to satisfy user information needs. Traditional IR systems, while effective in domains like web search, are constrained by their reliance on static, pre-defined information items. To this end, this paper introduces agentic information retrieval (Agentic IR), a transformative next-generation paradigm for IR driven by large language models (LLMs) and AI agents. The central shift in agentic IR is the evolving definition of ``information'' from static, pre-defined information items to dynamic, context-dependent information states. Information state refers to a particular information context that the user is right in within a dynamic environment, encompassing not only the acquired information items but also real-time user preferences, contextual factors, and decision-making processes. In such a way, traditional information retrieval, focused on acquiring relevant information items based on user queries, can be naturally extended to achieving the target information state given the user instruction, which thereby defines the agentic information retrieval. We systematically discuss agentic IR from various aspects, i.e., task formulation, architecture, evaluation, case studies, as well as challenges and future prospects. We believe that the concept of agentic IR introduced in this paper not only broadens the scope of information retrieval research but also lays the foundation for a more adaptive, interactive, and intelligent next-generation IR paradigm.
Authors: John M. Carpenter, Andrea Corvill\'on, Nihar B. Shah
Abstract: The increasing volume of papers and proposals that undergo peer review emphasizes the pressing need for greater automation to effectively manage the growing scale. In this study, we present the deployment and evaluation of machine learning and optimization techniques to assign proposals to reviewers that were developed for the Atacama Large Millimeter/submillimeter Array (ALMA) during the Cycle 10 Call for Proposals issued in 2023. Using topic modeling algorithms, we identify the proposal topics and assess reviewers' expertise based on their previous ALMA proposal submissions. We then apply an adapted version of the assignment optimization algorithm from PeerReview4All (Stelmakh et al. 2021) to maximize the alignment between proposal topics and reviewer expertise. Our evaluation shows a significant improvement in matching reviewer expertise: the median similarity score between the proposal topic and reviewer expertise increased by 51 percentage points compared to the previous cycle, and the percentage of reviewers reporting expertise in their assigned proposals rose by 20 percentage points. Furthermore, the assignment process proved highly effective in that no proposals required reassignment due to significant mismatches, resulting in a savings of 3 to 5 days of manual effort.
Authors: Jingyang Qiao, Zhizhong Zhang, Xin Tan, Yanyun Qu, Shouhong Ding, Yuan Xie
Abstract: Continual Instruction Tuning (CIT) is adopted to continually instruct Large Models to follow human intent data by data. It is observed that existing gradient update would heavily destroy the performance on previous datasets during CIT process. Instead, Exponential Moving Average (EMA), owns the ability to trace previous parameters, which can aid in decreasing forgetting. Nonetheless, its stable balance weight fails to deal with the ever-changing datasets, leading to the out-of-balance between plasticity and stability. In this paper, we propose a general continual instruction tuning framework to address the challenge. Starting from the trade-off prerequisite and EMA update, we propose the plasticity and stability ideal condition. Based on Taylor expansion in the loss function, we find the optimal balance weight can be automatically determined by the gradients and learned parameters. Therefore, we propose a stable-plasticity balanced coefficient to avoid knowledge confusion. Based on the semantic similarity of the instructions, we can determine whether to retrain or expand the training parameters and allocate the most suitable parameters for the testing instances. Extensive experiments across multiple continual instruction tuning benchmarks demonstrate that our approach not only enhances anti-forgetting capabilities but also significantly improves overall continual tuning performance. For example, based on LLaVA-7B, the forgetting is reduced from 5.42 to 1.93. Our code will be made publicly available soon.
Authors: Hyejun Jeong, Shiqing Ma, Amir Houmansadr
Abstract: Bias in machine learning models, particularly in Large Language Models, is a critical issue as these systems shape important societal decisions. While previous studies have examined bias in individual LLMs, comparisons of bias across models remain underexplored. To address this gap, we analyze 13 LLMs from five families, evaluating bias through output distribution across multiple dimensions using two datasets (4K and 1M questions). Our results show that fine-tuning has minimal impact on output distributions, and proprietary models tend to overly response as unknowns to minimize bias, compromising accuracy and utility. In addition, open-source models like Llama3-Chat and Gemma2-it demonstrate fairness comparable to proprietary models like GPT-4, challenging the assumption that larger, closed-source models are inherently less biased. We also find that bias scores for disambiguated questions are more extreme, raising concerns about reverse discrimination. These findings highlight the need for improved bias mitigation strategies and more comprehensive evaluation metrics for fairness in LLMs.
Authors: Junyu Luo, Xiao Luo, Xiusi Chen, Zhiping Xiao, Wei Ju, Ming Zhang
Abstract: Supervised fine-tuning (SFT) is crucial in adapting large language model (LLMs) to a specific domain or task. However, only a limited amount of labeled data is available in practical applications, which poses a severe challenge for SFT in yielding satisfactory results. Therefore, a data-efficient framework that can fully exploit labeled and unlabeled data for LLM fine-tuning is highly anticipated.Towards this end, we introduce a semi-supervised fine-tuning(SemiFT) task and a framework named SemiEvol for LLM alignment from a propagate-and-select manner. For knowledge propagation, SemiEvol adopts a bi-level approach, propagating knowledge from labeled data to unlabeled data through both in-weight and in-context methods. For knowledge selection, SemiEvol incorporates a collaborative learning mechanism, selecting higher-quality pseudo-response samples. We conducted experiments using GPT-4o-mini and Llama-3.1 on seven general or domain-specific datasets, demonstrating significant improvements in model performance on target data. Furthermore, we compared SemiEvol with SFT and self-evolution methods, highlighting its practicality in hybrid data scenarios.
Authors: Bryan R. Christ, Zack Gottesman, Jonathan Kropko, Thomas Hartvigsen
Abstract: Math reasoning is an active area of Large Language Model (LLM) research because it is a hallmark of artificial intelligence and has implications in several domains, including math education. However, few works have explored how math reasoning is encoded within LLM parameters and if it is a skill that can be isolated within models. Doing so could allow targeted intervention to improve math performance without altering non-math behavior and foster understanding of how models encode math reasoning. We introduce Math Neurosurgery (MathNeuro), a computationally efficient method we use to isolate math-specific parameters in LLMs using only forward passes. MathNeuro builds on existing work by using weights and activations to calculate parameter importance, but isolates math-specific parameters by filtering out those important for general language tasks. Through pruning parameters MathNeuro identifies, we delete a LLM's math reasoning ability without significantly impacting its general language ability. Scaling the identified parameters by a small constant improves a pretrained or instruction-tuned LLM's performance by 4-17% on GSM8K and 5-35% on MATH while leaving non-math behavior unaltered. MathNeuro is also data efficient: most of its effectiveness holds when identifying math-specific parameters using a single sample. MathNeuro highlights the potential for future work to intervene on math-specific parameters.
Authors: Elisabeth Kirsten, Ivan Habernal, Vedant Nanda, Muhammad Bilal Zafar
Abstract: Last few years have seen unprecedented advances in capabilities of Large Language Models (LLMs). These advancements promise to benefit a vast array of application domains. However, due to their immense size, performing inference with LLMs is both costly and slow. Consequently, a plethora of recent work has proposed strategies to enhance inference efficiency, e.g., quantization, pruning, and caching. These acceleration strategies reduce the inference cost and latency, often by several factors, while maintaining much of the predictive performance measured via common benchmarks. In this work, we explore another critical aspect of LLM performance: demographic bias in model generations due to inference acceleration optimizations. Using a wide range of metrics, we probe bias in model outputs from a number of angles. Analysis of outputs before and after inference acceleration shows significant change in bias. Worryingly, these bias effects are complex and unpredictable. A combination of an acceleration strategy and bias type may show little bias change in one model but may lead to a large effect in another. Our results highlight a need for in-depth and case-by-case evaluation of model bias after it has been modified to accelerate inference.
Authors: Seonil Son, Ju-Min Oh, Heegon Jin, Cheolhun Jang, Jeongbeom Jeong, Kuntae Kim
Abstract: Most existing benchmarking approaches for evaluating the output quality of large language models (LLMs) rely on comparing LLM responses to predefined references. Such methods, based on static datasets, quickly become outdated as LLM capabilities and use cases evolve. In this work, we introduce VARCO Arena--a novel, cost-effective, and robust benchmarking approach that leverages a single-elimination tournament structure to minimize the number of required comparisons while eliminating the need for static references or costly human annotations. We validate our approach through two experiments: (i) a simulation study that examines its robustness under various conditions, and (ii) an empirical evaluation using publicly available benchmark prompts. In both experiments, VARCO Arena consistently outperforms current LLM benchmarking practices, achieving stronger correlations with human-established Elo ratings. Our results demonstrate that VARCO Arena not only produces reliable LLM rankings but also provides a scalable, adaptable solution for qualitative evaluation across diverse, customized use cases.
Authors: Aliyah R. Hsu, James Zhu, Zhichao Wang, Bin Bi, Shubham Mehrotra, Shiva K. Pentyala, Katherine Tan, Xiang-Bo Mao, Roshanak Omrani, Sougata Chaudhuri, Regunathan Radhakrishnan, Sitaram Asur, Claire Na Cheng, Bin Yu
Abstract: LLMs have demonstrated impressive proficiency in generating coherent and high-quality text, making them valuable across a range of text-generation tasks. However, rigorous evaluation of this generated content is crucial, as ensuring its quality remains a significant challenge due to persistent issues such as factual inaccuracies and hallucination. This paper introduces three fine-tuned general-purpose LLM autoevaluators, REC-8B, REC-12B and REC-70B, specifically designed to evaluate generated text across several dimensions: faithfulness, instruction following, coherence, and completeness. These models not only provide ratings for these metrics but also offer detailed explanation and verifiable citation, thereby enhancing trust in the content. Moreover, the models support various citation modes, accommodating different requirements for latency and granularity. Extensive evaluations on diverse benchmarks demonstrate that our general-purpose LLM auto-evaluator, REC-70B, outperforms state-of-the-art LLMs, excelling in content evaluation by delivering better quality explanation and citation with minimal bias. It achieves Rank #1 as of Feb 15th, 2025 as a generative model on the RewardBench leaderboard under the model name TextEval-Llama3.1-70B. Our REC dataset and models are available at https://github.com/adelaidehsu/REC.
Authors: Yansong Qu, Zixuan Xu, Zilin Huang, Zihao Sheng, Tiantian Chen, Sikai Chen
Abstract: Semantic scene completion (SSC) is essential for achieving comprehensive perception in autonomous driving systems. However, existing SSC methods often overlook the high deployment costs in real-world applications. Traditional architectures, such as 3D Convolutional Neural Networks (3D CNNs) and self-attention mechanisms, face challenges in efficiently capturing long-range dependencies within 3D voxel grids, limiting their effectiveness. To address these issues, we introduce MetaSSC, a novel meta-learning-based framework for SSC that leverages deformable convolution, large-kernel attention, and the Mamba (D-LKA-M) model. Our approach begins with a voxel-based semantic segmentation (SS) pretraining task, aimed at exploring the semantics and geometry of incomplete regions while acquiring transferable meta-knowledge. Using simulated cooperative perception datasets, we supervise the perception training of a single vehicle using aggregated sensor data from multiple nearby connected autonomous vehicles (CAVs), generating richer and more comprehensive labels. This meta-knowledge is then adapted to the target domain through a dual-phase training strategy that does not add extra model parameters, enabling efficient deployment. To further enhance the model's capability in capturing long-sequence relationships within 3D voxel grids, we integrate Mamba blocks with deformable convolution and large-kernel attention into the backbone network. Extensive experiments demonstrate that MetaSSC achieves state-of-the-art performance, significantly outperforming competing models while also reducing deployment costs.
Authors: Brian Hsuan-Cheng Liao, Yingjie Xu, Chih-Hong Cheng, Hasan Esen, Alois Knoll
Abstract: This paper presents a novel monitoring framework that infers the level of collision risk for autonomous vehicles (AVs) based on their object detection performance. The framework takes two sets of predictions from different algorithms and associates their inconsistencies with the collision risk via fuzzy inference. The first set of predictions is obtained by retrieving safety-critical 2.5D objects from a depth map, and the second set comes from the ordinary AV's 3D object detector. We experimentally validate that, based on Intersection-over-Union (IoU) and a depth discrepancy measure, the inconsistencies between the two sets of predictions strongly correlate to the error of the 3D object detector against ground truths. This correlation allows us to construct a fuzzy inference system and map the inconsistency measures to an AV collision risk indicator. In particular, we optimize the fuzzy inference system towards an existing offline metric that matches AV collision rates well. Lastly, we validate our monitor's capability to produce relevant risk estimates with the large-scale nuScenes dataset and demonstrate that it can safeguard an AV in closed-loop simulations.
Authors: Wadhah Zai El Amri, Malte Kuhlmann, Nicol\'as Navarro-Guerrero
Abstract: Tactile perception is essential for human interaction with the environment and is becoming increasingly crucial in robotics. Tactile sensors like the BioTac mimic human fingertips and provide detailed interaction data. Despite its utility in applications like slip detection and object identification, this sensor is now deprecated, making many valuable datasets obsolete. However, recreating similar datasets with newer sensor technologies is both tedious and time-consuming. Therefore, adapting these existing datasets for use with new setups and modalities is crucial. In response, we introduce ACROSS, a novel framework for translating data between tactile sensors by exploiting sensor deformation information. We demonstrate the approach by translating BioTac signals into the DIGIT sensor. Our framework consists of first converting the input signals into 3D deformation meshes. We then transition from the 3D deformation mesh of one sensor to the mesh of another, and finally convert the generated 3D deformation mesh into the corresponding output space. We demonstrate our approach to the most challenging problem of going from a low-dimensional tactile representation to a high-dimensional one. In particular, we transfer the tactile signals of a BioTac sensor to DIGIT tactile images. Our approach enables the continued use of valuable datasets and data exchange between groups with different setups.
Authors: Xue Tan, Hao Luan, Mingyu Luo, Xiaoyan Sun, Ping Chen, Jun Dai
Abstract: Retrieval-Augmented Generation (RAG) enriches the input to LLMs by retrieving information from the relevant knowledge database, enabling them to produce responses that are more accurate and contextually appropriate. It is worth noting that the knowledge database, being sourced from publicly available channels such as Wikipedia, inevitably introduces a new attack surface. RAG poisoning involves injecting malicious texts into the knowledge database, ultimately leading to the generation of the attacker's target response (also called poisoned response). However, there are currently limited methods available for detecting such poisoning attacks. We aim to bridge the gap in this work. Particularly, we introduce RevPRAG, a flexible and automated detection pipeline that leverages the activations of LLMs for poisoned response detection. Our investigation uncovers distinct patterns in LLMs' activations when generating correct responses versus poisoned responses. Our results on multiple benchmark datasets and RAG architectures show our approach could achieve 98% true positive rate, while maintaining false positive rates close to 1%.
Authors: Massine El Khader, Elias Al Bouzidi, Abdellah Oumida, Mohammed Sbaihi, Eliott Binard, Jean-Philippe Poli, Wassila Ouerdane, Boussad Addad, Katarzyna Kapusta
Abstract: Recent advances in Diffusion Models have enabled the generation of images from text, with powerful closed-source models like DALL-E and Midjourney leading the way. However, open-source alternatives, such as StabilityAI's Stable Diffusion, offer comparable capabilities. These open-source models, hosted on Hugging Face, come equipped with ethical filter protections designed to prevent the generation of explicit images. This paper reveals first their limitations and then presents a novel text-based safety filter that outperforms existing solutions. Our research is driven by the critical need to address the misuse of AI-generated content, especially in the context of information warfare. DiffGuard enhances filtering efficacy, achieving a performance that surpasses the best existing filters by over 14%.
Authors: Pengjie Zhou, Haoyu Wei, Huiming Zhang
Abstract: Reinforcement Learning (RL) is a widely researched area in artificial intelligence that focuses on teaching agents decision-making through interactions with their environment. A key subset includes stochastic multi-armed bandit (MAB) and continuum-armed bandit (SCAB) problems, which model sequential decision-making under uncertainty. This review outlines the foundational models and assumptions of bandit problems, explores non-asymptotic theoretical tools like concentration inequalities and minimax regret bounds, and compares frequentist and Bayesian algorithms for managing exploration-exploitation trade-offs. Additionally, we explore K-armed contextual bandits and SCAB, focusing on their methodologies and regret analyses. We also examine the connections between SCAB problems and functional data analysis. Finally, we highlight recent advances and ongoing challenges in the field.
Authors: Adit Gupta, Momin Siddiqui, Glen Smith, Jenn Reddig, Christopher MacLellan
Abstract: This work examines the sociotechnical factors that influence the adoption and usage of intelligent tutoring systems in self-directed learning contexts, focusing specifically on adult learners. The study is divided into two parts. First, we present Apprentice Tutors, a novel intelligent tutoring system designed to address the unique needs of adult learners. The platform includes adaptive problem selection, real-time feedback, and visual dashboards to support learning in college algebra topics. Second, we investigate the specific needs and experiences of adult users through a deployment study and a series of focus groups. Using thematic analysis, we identify key challenges and opportunities to improve tutor design and adoption. Based on these findings, we offer actionable design recommendations to help developers create intelligent tutoring systems that better align with the motivations and learning preferences of adult learners. This work contributes to a wider understanding of how to improve educational technologies to support lifelong learning and professional development.
Authors: Chaozheng Wen, Jingwen Tong, Yingdong Hu, Zehong Lin, Jun Zhang
Abstract: Wireless channel modeling plays a pivotal role in designing, analyzing, and optimizing wireless communication systems. Nevertheless, developing an effective channel modeling approach has been a longstanding challenge. This issue has been escalated due to the denser network deployment, larger antenna arrays, and wider bandwidth in 5G and beyond networks. To address this challenge, we put forth WRF-GS, a novel framework for channel modeling based on wireless radiation field (WRF) reconstruction using 3D Gaussian splatting. WRF-GS employs 3D Gaussian primitives and neural networks to capture the interactions between the environment and radio signals, enabling efficient WRF reconstruction and visualization of the propagation characteristics. The reconstructed WRF can then be used to synthesize the spatial spectrum for comprehensive wireless channel characterization. Notably, with a small number of measurements, WRF-GS can synthesize new spatial spectra within milliseconds for a given scene, thereby enabling latency-sensitive applications. Experimental results demonstrate that WRF-GS outperforms existing methods for spatial spectrum synthesis, such as ray tracing and other deep-learning approaches. Moreover, WRF-GS achieves superior performance in the channel state information prediction task, surpassing existing methods by a significant margin of more than 2.43 dB.
Authors: Minzheng Wang, Xinghua Zhang, Kun Chen, Nan Xu, Haiyang Yu, Fei Huang, Wenji Mao, Yongbin Li
Abstract: Large language models (LLMs) enabled dialogue systems have become one of the central modes in human-machine interaction, which bring about vast amounts of conversation logs and increasing demand for dialogue generation. The dialogue's life-cycle spans from $\textit{Prelude}$ through $\textit{Interlocution}$ to $\textit{Epilogue}$, encompassing rich dialogue elements. Despite large volumes of dialogue-related studies, there is a lack of systematic investigation into the dialogue stages to frame benchmark construction that covers comprehensive dialogue elements. This hinders the precise modeling, generation and assessment of LLMs-based dialogue systems. To bridge this gap, in this paper, we introduce a new research task--$\textbf{D}$ialogue $\textbf{E}$lement $\textbf{MO}$deling, including $\textit{Element Awareness}$ and $\textit{Dialogue Agent Interaction}$, and propose a novel benchmark, $\textbf{DEMO}$, designed for a comprehensive dialogue modeling and assessment. On this basis, we further build the DEMO agent with the adept ability to model dialogue elements via imitation learning. Extensive experiments on DEMO indicate that current representative LLMs still have considerable potential for enhancement, and our DEMO agent performs well in both dialogue element modeling and out-of-domain tasks.
Authors: Bo Lv, Chen Tang, Yanan Zhang, Xin Liu, Yue Yu, Ping Luo
Abstract: Ensembles of generative large language models (LLMs) can integrate the strengths of different LLMs to compensate for the limitations of individual models. However, recent work has focused on training an additional fusion model to combine complete responses from multiple LLMs, failing to tap into their collaborative potential to generate higher-quality responses. Moreover, as the additional fusion model is trained on a specialized dataset, these methods struggle with generalizing to open-domain queries from online users. In this paper, we propose SpecFuse, a novel ensemble framework that outputs the fused result by iteratively producing the next segment through collaboration among LLMs. This is achieved through cyclic execution of its inference and verification components. In each round, the inference component invokes each base LLM to generate candidate segments in parallel, and the verify component calls these LLMs again to predict the ranking of the segments. The top-ranked segment is then broadcast to all LLMs, encouraging them to generate higher-quality segments in the next round. This approach also allows the base LLMs to be plug-and-play, without any training or adaptation, avoiding generalization limitations. Furthermore, to conserve computational resources, we propose a model exit mechanism that dynamically excludes models exhibiting poor performance in previous rounds during each query response. In this way, it effectively reduces the number of model calls while maintaining overall performance.
Authors: Wonjin Lee, Kyumin Kim, Sungjae Lee, Jihun Lee, Kwang In Kim
Abstract: Applying language models (LMs) to tables is challenging due to the inherent structural differences between two-dimensional tables and one-dimensional text for which the LMs were originally designed. Furthermore, when applying linearized tables to LMs, the maximum token lengths often imposed in self-attention calculations make it difficult to comprehensively understand the context spread across large tables. To address these challenges, we present PieTa (Piece of Table), a new framework for subtable-based question answering (QA). PieTa operates through an iterative process of dividing tables into smaller windows, using LMs to select relevant cells within each window, and merging these cells into a subtable. This multi-resolution approach captures dependencies across multiple rows and columns while avoiding the limitations caused by long context inputs. Instantiated as a simple iterative subtable union algorithm, PieTa demonstrates improved performance over previous subtable-based QA approaches.
Authors: Guangsheng Bao, Yanbin Zhao, Juncai He, Yue Zhang
Abstract: Advanced large language models (LLMs) can generate text almost indistinguishable from human-written text, highlighting the importance of LLM-generated text detection. However, current zero-shot techniques face challenges as white-box methods are restricted to use weaker open-source LLMs, and black-box methods are limited by partial observation from stronger proprietary LLMs. It seems impossible to enable white-box methods to use proprietary models because API-level access to the models neither provides full predictive distributions nor inner embeddings. To traverse the divide, we propose **Glimpse**, a probability distribution estimation approach, predicting the full distributions from partial observations. Despite the simplicity of Glimpse, we successfully extend white-box methods like Entropy, Rank, Log-Rank, and Fast-DetectGPT to latest proprietary models. Experiments show that Glimpse with Fast-DetectGPT and GPT-3.5 achieves an average AUROC of about 0.95 in five latest source models, improving the score by 51% relative to the remaining space of the open source baseline. It demonstrates that the latest LLMs can effectively detect their own outputs, suggesting that advanced LLMs may be the best shield against themselves. We release our code and data at https://github.com/baoguangsheng/glimpse.
Authors: Yu Yan, Sheng Sun, Junqi Tong, Min Liu, Qi Li
Abstract: Metaphor serves as an implicit approach to convey information, while enabling the generalized comprehension of complex subjects. However, metaphor can potentially be exploited to bypass the safety alignment mechanisms of Large Language Models (LLMs), leading to the theft of harmful knowledge. In our study, we introduce a novel attack framework that exploits the imaginative capacity of LLMs to achieve jailbreaking, the J\underline{\textbf{A}}ilbreak \underline{\textbf{V}}ia \underline{\textbf{A}}dversarial Me\underline{\textbf{TA}} -pho\underline{\textbf{R}} (\textit{AVATAR}). Specifically, to elicit the harmful response, AVATAR extracts harmful entities from a given harmful target and maps them to innocuous adversarial entities based on LLM's imagination. Then, according to these metaphors, the harmful target is nested within human-like interaction for jailbreaking adaptively. Experimental results demonstrate that AVATAR can effectively and transferablly jailbreak LLMs and achieve a state-of-the-art attack success rate across multiple advanced LLMs. Our study exposes a security risk in LLMs from their endogenous imaginative capabilities. Furthermore, the analytical study reveals the vulnerability of LLM to adversarial metaphors and the necessity of developing defense methods against jailbreaking caused by the adversarial metaphor. \textcolor{orange}{ \textbf{Warning: This paper contains potentially harmful content from LLMs.}}
Authors: Jiayuan Su, Zhaopeng Feng, Jiamei Zheng, Jiahan Ren, Yan Zhang, Jian Wu, Hongwei Wang, Zuozhu Liu
Abstract: Recent advancements in large language models (LLMs) have given rise to the LLM-as-a-judge paradigm, showcasing their potential to deliver human-like judgments. However, in the field of machine translation (MT) evaluation, current LLM-as-a-judge methods fall short of learned automatic metrics. In this paper, we propose Multidimensional Multi-Agent Debate (M-MAD), a systematic LLM-based multi-agent framework for advanced LLM-as-a-judge MT evaluation. Our findings demonstrate that M-MAD achieves significant advancements by (1) decoupling heuristic MQM criteria into distinct evaluation dimensions for fine-grained assessments; (2) employing multi-agent debates to harness the collaborative reasoning capabilities of LLMs; (3) synthesizing dimension-specific results into a final evaluation judgment to ensure robust and reliable outcomes. Comprehensive experiments show that M-MAD not only outperforms all existing LLM-as-a-judge methods but also competes with state-of-the-art reference-based automatic metrics, even when powered by a suboptimal model like GPT-4o mini. Detailed ablations and analysis highlight the superiority of our framework design, offering a fresh perspective for LLM-as-a-judge paradigm. Our code and data are publicly available at https://github.com/SU-JIAYUAN/M-MAD.
Authors: Zijie Chen, Zhanchao Zhou, Yu Lu, Renjun Xu, Lili Pan, Zhenzhong Lan
Abstract: Solving NP-hard problems traditionally relies on heuristics, yet manually designing effective heuristics for complex problems remains a significant challenge. While recent advancements like FunSearch have shown that large language models (LLMs) can be integrated into evolutionary algorithms (EAs) for heuristic design, their potential is hindered by limitations in balancing exploitation and exploration. We introduce Quality-Uncertainty Balanced Evolution (QUBE), a novel approach that enhances LLM+EA methods by redefining the priority criterion within the FunSearch framework. QUBE employs the Quality-Uncertainty Trade-off Criterion (QUTC), based on our proposed Uncertainty-Inclusive Quality metric, to evaluate and guide the evolutionary process. Through extensive experiments on challenging NP-complete problems, QUBE demonstrates significant performance improvements over FunSearch and baseline methods. Our code are available at https://github.com/zzjchen/QUBE\_code.
Authors: Zhen Li, Yupeng Su, Runming Yang, Congkai Xie, Zheng Wang, Zhongwei Xie, Ngai Wong, Hongxia Yang
Abstract: Large language models have achieved significant advancements in complex mathematical reasoning benchmarks, such as MATH. However, their substantial computational requirements present challenges for practical deployment. Model quantization has emerged as an effective strategy to reduce memory usage and computational costs by employing lower precision and bit-width representations. In this study, we systematically evaluate the impact of quantization on mathematical reasoning tasks. Our results demonstrate that aggressive quantization methods like AWQ and GPTQ introduce up to 32.39% accuracy degradation (average 11.31%) on Llama-3 models, particularly in numerical computation and reasoning planning. To address this, we introduce a multidimensional evaluation framework combining qualitative capability analysis and quantitative error assessment. We further develop targeted recovery strategies, showing that fine-tuning quantized models on only 545 task-specific examples for 3 minutes on 4 GPUs effectively restores reasoning capabilities to near full-precision levels. Additionally, our error assessment pipeline achieves 98.9% accuracy in diagnosing and localizing errors across 3,366 failure cases, providing actionable insights for mitigating quantization-induced degradation.
Authors: Fabian David Schmidt, Ivan Vuli\'c, Goran Glava\v{s}, David Ifeoluwa Adelani
Abstract: Spoken language understanding (SLU) is indispensable for half of all living languages that lack a formal writing system, since these languages cannot pair automatic speech recognition (ASR) with language models to benefit from language technology. Even if low-resource languages possess a writing system, ASR for these languages remains unreliable due to limited bimodal speech and text training data. Better SLU can strengthen the robustness of massively multilingual ASR by levering language semantics to disambiguate utterances via context or exploiting semantic similarities across languages. However, the evaluation of multilingual SLU remains limited to shallow tasks such as intent classification or language identification. To address this, we present Fleurs-SLU, a multilingual SLU benchmark that encompasses (i) 692 hours of speech for topical utterance classification in 102 languages and (ii) multiple-choice question answering through listening comprehension spanning 944 hours of speech across 92 languages. We extensively evaluate both end-to-end speech classification models and cascaded systems that combine speech-to-text transcription with subsequent classification by large language models on Fleurs-SLU. Our results show that cascaded systems exhibit greater robustness in multilingual SLU tasks, though speech encoders can achieve competitive performance in topical speech classification when appropriately pre-trained. We further find a strong correlation between robust multilingual ASR, effective speech-to-text translation, and strong multilingual SLU, highlighting the mutual benefits between acoustic and semantic speech representations.
Authors: Luqi Zhang, Haiping Wang, Chong Liu, Zhen Dong, Bisheng Yang
Abstract: The point clouds collected by the Airborne Laser Scanning (ALS) system provide accurate 3D information of urban land covers. By utilizing multi-temporal ALS point clouds, semantic changes in urban area can be captured, demonstrating significant potential in urban planning, emergency management, and infrastructure maintenance. Existing 3D change detection methods struggle to efficiently extract multi-class semantic information and change features, still facing the following challenges: (1) the difficulty of accurately modeling cross-temporal point clouds spatial relationships for effective change feature extraction; (2) class imbalance of change samples which hinders distinguishability of semantic features; (3) the lack of real-world datasets for 3D semantic change detection. To resolve these challenges, we propose the Multi-task Enhanced Cross-temporal Point Transformer (ME-CPT) network. ME-CPT establishes spatiotemporal correspondences between point cloud across different epochs and employs attention mechanisms to jointly extract semantic change features, facilitating information exchange and change comparison. Additionally, we incorporate a semantic segmentation task and through the multi-task training strategy, further enhance the distinguishability of semantic features, reducing the impact of class imbalance in change types. Moreover, we release a 22.5 $km^2$ 3D semantic change detection dataset, offering diverse scenes for comprehensive evaluation. Experiments on multiple datasets show that the proposed MT-CPT achieves superior performance compared to existing state-of-the-art methods. The source code and dataset will be released upon acceptance at https://github.com/zhangluqi0209/ME-CPT.
Authors: Long Phan (Michael Pokorny), Alice Gatti (Michael Pokorny), Ziwen Han (Michael Pokorny), Nathaniel Li (Michael Pokorny), Josephina Hu (Michael Pokorny), Hugh Zhang (Michael Pokorny), Chen Bo Calvin Zhang (Michael Pokorny), Mohamed Shaaban (Michael Pokorny), John Ling (Michael Pokorny), Sean Shi (Michael Pokorny), Michael Choi (Michael Pokorny), Anish Agrawal (Michael Pokorny), Arnav Chopra (Michael Pokorny), Adam Khoja (Michael Pokorny), Ryan Kim (Michael Pokorny), Richard Ren (Michael Pokorny), Jason Hausenloy (Michael Pokorny), Oliver Zhang (Michael Pokorny), Mantas Mazeika (Michael Pokorny), Tung Nguyen (Michael Pokorny), Daron Anderson (Michael Pokorny), Imad Ali Shah (Michael Pokorny), Mikhail Doroshenko (Michael Pokorny), Alun Cennyth Stokes (Michael Pokorny), Mobeen Mahmood (Michael Pokorny), Jaeho Lee (Michael Pokorny), Oleksandr Pokutnyi (Michael Pokorny), Oleg Iskra (Michael Pokorny), Jessica P. Wang (Michael Pokorny), Robert Gerbicz (Michael Pokorny), John-Clark Levin (Michael Pokorny), Serguei Popov (Michael Pokorny), Fiona Feng (Michael Pokorny), Steven Y. Feng (Michael Pokorny), Haoran Zhao (Michael Pokorny), Michael Yu (Michael Pokorny), Varun Gangal (Michael Pokorny), Chelsea Zou (Michael Pokorny), Zihan Wang (Michael Pokorny), Mstyslav Kazakov (Michael Pokorny), Geoff Galgon (Michael Pokorny), Johannes Schmitt (Michael Pokorny), Alvaro Sanchez (Michael Pokorny), Yongki Lee (Michael Pokorny), Will Yeadon (Michael Pokorny), Scott Sauers (Michael Pokorny), Marc Roth (Michael Pokorny), Chidozie Agu (Michael Pokorny), S{\o}ren Riis (Michael Pokorny), Fabian Giska (Michael Pokorny), Saiteja Utpala (Michael Pokorny), Antrell Cheatom (Michael Pokorny), Zachary Giboney (Michael Pokorny), Gashaw M. Goshu (Michael Pokorny), Sarah-Jane Crowson (Michael Pokorny), Mohinder Maheshbhai Naiya (Michael Pokorny), Noah Burns (Michael Pokorny), Lennart Finke (Michael Pokorny), Zerui Cheng (Michael Pokorny), Hyunwoo Park (Michael Pokorny), Francesco Fournier-Facio (Michael Pokorny), Jennifer Zampese (Michael Pokorny), John Wydallis (Michael Pokorny), John B. Wydallis (Michael Pokorny), Ryan G. Hoerr (Michael Pokorny), Mark Nandor (Michael Pokorny), Tim Gehrunger (Michael Pokorny), Jiaqi Cai (Michael Pokorny), Ben McCarty (Michael Pokorny), Jungbae Nam (Michael Pokorny), Edwin Taylor (Michael Pokorny), Jun Jin (Michael Pokorny), Gautier Abou Loume (Michael Pokorny), Hangrui Cao (Michael Pokorny), Alexis C Garretson (Michael Pokorny), Damien Sileo (Michael Pokorny), Qiuyu Ren (Michael Pokorny), Doru Cojoc (Michael Pokorny), Pavel Arkhipov (Michael Pokorny), Usman Qazi (Michael Pokorny), Aras Bacho (Michael Pokorny), Lianghui Li (Michael Pokorny), Sumeet Motwani (Michael Pokorny), Christian Schroeder de Witt (Michael Pokorny), Alexei Kopylov (Michael Pokorny), Johannes Veith (Michael Pokorny), Eric Singer (Michael Pokorny), Paolo Rissone (Michael Pokorny), Jaehyeok Jin (Michael Pokorny), Jack Wei Lun Shi (Michael Pokorny), Chris G. Willcocks (Michael Pokorny), Ameya Prabhu (Michael Pokorny), Longke Tang (Michael Pokorny), Kevin Zhou (Michael Pokorny), Emily de Oliveira Santos (Michael Pokorny), Andrey Pupasov Maksimov (Michael Pokorny), Edward Vendrow (Michael Pokorny), Kengo Zenitani (Michael Pokorny), Joshua Robinson (Michael Pokorny), Aleksandar Mikov (Michael Pokorny), Julien Guillod (Michael Pokorny), Yuqi Li (Michael Pokorny), Ben Pageler (Michael Pokorny), Joshua Vendrow (Michael Pokorny), Vladyslav Kuchkin (Michael Pokorny), Pierre Marion (Michael Pokorny), Denis Efremov (Michael Pokorny), Jayson Lynch (Michael Pokorny), Kaiqu Liang (Michael Pokorny), Andrew Gritsevskiy (Michael Pokorny), Dakotah Martinez (Michael Pokorny), Nick Crispino (Michael Pokorny), Dimitri Zvonkine (Michael Pokorny), Natanael Wildner Fraga (Michael Pokorny), Saeed Soori (Michael Pokorny), Ori Press (Michael Pokorny), Henry Tang (Michael Pokorny), Julian Salazar (Michael Pokorny), Sean R. Green (Michael Pokorny), Lina Br\"ussel (Michael Pokorny), Moon Twayana (Michael Pokorny), Aymeric Dieuleveut (Michael Pokorny), T. Ryan Rogers (Michael Pokorny), Wenjin Zhang (Michael Pokorny), Ross Finocchio (Michael Pokorny), Bikun Li (Michael Pokorny), Jinzhou Yang (Michael Pokorny), Arun Rao (Michael Pokorny), Gabriel Loiseau (Michael Pokorny), Mikhail Kalinin (Michael Pokorny), Marco Lukas (Michael Pokorny), Ciprian Manolescu (Michael Pokorny), Nate Stambaugh (Michael Pokorny), Subrata Mishra (Michael Pokorny), Ariel Ghislain Kemogne Kamdoum (Michael Pokorny), Tad Hogg (Michael Pokorny), Alvin Jin (Michael Pokorny), Carlo Bosio (Michael Pokorny), Gongbo Sun (Michael Pokorny), Brian P Coppola (Michael Pokorny), Haline Heidinger (Michael Pokorny), Rafael Sayous (Michael Pokorny), Stefan Ivanov (Michael Pokorny), Joseph M Cavanagh (Michael Pokorny), Jiawei Shen (Michael Pokorny), Joseph Marvin Imperial (Michael Pokorny), Philippe Schwaller (Michael Pokorny), Shaipranesh Senthilkuma (Michael Pokorny), Andres M Bran (Michael Pokorny), Andres Algaba (Michael Pokorny), Brecht Verbeken (Michael Pokorny), Kelsey Van den Houte (Michael Pokorny), Lynn Van Der Sypt (Michael Pokorny), David Noever (Michael Pokorny), Lisa Schut (Michael Pokorny), Ilia Sucholutsky (Michael Pokorny), Evgenii Zheltonozhskii (Michael Pokorny), Qiaochu Yuan (Michael Pokorny), Derek Lim (Michael Pokorny), Richard Stanley (Michael Pokorny), Shankar Sivarajan (Michael Pokorny), Tong Yang (Michael Pokorny), John Maar (Michael Pokorny), Julian Wykowski (Michael Pokorny), Mart\'i Oller (Michael Pokorny), Jennifer Sandlin (Michael Pokorny), Anmol Sahu (Michael Pokorny), Cesare Giulio Ardito (Michael Pokorny), Yuzheng Hu (Michael Pokorny), Felipe Meneguitti Dias (Michael Pokorny), Tobias Kreiman (Michael Pokorny), Kaivalya Rawal (Michael Pokorny), Tobias Garcia Vilchis (Michael Pokorny), Yuexuan Zu (Michael Pokorny), Martin Lackner (Michael Pokorny), James Koppel (Michael Pokorny), Jeremy Nguyen (Michael Pokorny), Daniil S. Antonenko (Michael Pokorny), Steffi Chern (Michael Pokorny), Bingchen Zhao (Michael Pokorny), Pierrot Arsene (Michael Pokorny), Sergey Ivanov (Michael Pokorny), Rafa{\l} Po\'swiata (Michael Pokorny), Chenguang Wang (Michael Pokorny), Daofeng Li (Michael Pokorny), Donato Crisostomi (Michael Pokorny), Ali Dehghan (Michael Pokorny), Andrea Achilleos (Michael Pokorny), John Arnold Ambay (Michael Pokorny), Benjamin Myklebust (Michael Pokorny), Archan Sen (Michael Pokorny), David Perrella (Michael Pokorny), Nurdin Kaparov (Michael Pokorny), Mark H Inlow (Michael Pokorny), Allen Zang (Michael Pokorny), Kalyan Ramakrishnan (Michael Pokorny), Daniil Orel (Michael Pokorny), Vladislav Poritski (Michael Pokorny), Shalev Ben-David (Michael Pokorny), Zachary Berger (Michael Pokorny), Parker Whitfill (Michael Pokorny), Michael Foster (Michael Pokorny), Daniel Munro (Michael Pokorny), Linh Ho (Michael Pokorny), Dan Bar Hava (Michael Pokorny), Aleksey Kuchkin (Michael Pokorny), Robert Lauff (Michael Pokorny), David Holmes (Michael Pokorny), Frank Sommerhage (Michael Pokorny), Anji Zhang (Michael Pokorny), Richard Moat (Michael Pokorny), Keith Schneider (Michael Pokorny), Daniel Pyda (Michael Pokorny), Zakayo Kazibwe (Michael Pokorny), Mukhwinder Singh (Michael Pokorny), Don Clarke (Michael Pokorny), Dae Hyun Kim (Michael Pokorny), Sara Fish (Michael Pokorny), Veit Elser (Michael Pokorny), Victor Efren Guadarrama Vilchis (Michael Pokorny), Immo Klose (Michael Pokorny), Christoph Demian (Michael Pokorny), Ujjwala Anantheswaran (Michael Pokorny), Adam Zweiger (Michael Pokorny), Guglielmo Albani (Michael Pokorny), Jeffery Li (Michael Pokorny), Nicolas Daans (Michael Pokorny), Maksim Radionov (Michael Pokorny), V\'aclav Rozho\v{n} (Michael Pokorny), Vincent Ginis (Michael Pokorny), Ziqiao Ma (Michael Pokorny), Christian Stump (Michael Pokorny), Jacob Platnick (Michael Pokorny), Volodymyr Nevirkovets (Michael Pokorny), Luke Basler (Michael Pokorny), Marco Piccardo (Michael Pokorny), Niv Cohen (Michael Pokorny), Virendra Singh (Michael Pokorny), Josef Tkadlec (Michael Pokorny), Paul Rosu (Michael Pokorny), Alan Goldfarb (Michael Pokorny), Piotr Padlewski (Michael Pokorny), Stanislaw Barzowski (Michael Pokorny), Kyle Montgomery (Michael Pokorny), Aline Menezes (Michael Pokorny), Arkil Patel (Michael Pokorny), Zixuan Wang (Michael Pokorny), Jamie Tucker-Foltz (Michael Pokorny), Jack Stade (Michael Pokorny), Declan Grabb (Michael Pokorny), Tom Goertzen (Michael Pokorny), Fereshteh Kazemi (Michael Pokorny), Jeremiah Milbauer (Michael Pokorny), Abhishek Shukla (Michael Pokorny), Hossam Elgnainy (Michael Pokorny), Yan Carlos Leyva Labrador (Michael Pokorny), Hao He (Michael Pokorny), Ling Zhang (Michael Pokorny), Alan Givr\'e (Michael Pokorny), Hew Wolff (Michael Pokorny), G\"ozdenur Demir (Michael Pokorny), Muhammad Fayez Aziz (Michael Pokorny), Younesse Kaddar (Michael Pokorny), Ivar \"Angquist (Michael Pokorny), Yanxu Chen (Michael Pokorny), Elliott Thornley (Michael Pokorny), Robin Zhang (Michael Pokorny), Jiayi Pan (Michael Pokorny), Antonio Terpin (Michael Pokorny), Niklas Muennighoff (Michael Pokorny), Hailey Schoelkopf (Michael Pokorny), Eric Zheng (Michael Pokorny), Avishy Carmi (Michael Pokorny), Jainam Shah (Michael Pokorny), Ethan D. L. Brown (Michael Pokorny), Kelin Zhu (Michael Pokorny), Max Bartolo (Michael Pokorny), Richard Wheeler (Michael Pokorny), Andrew Ho (Michael Pokorny), Shaul Barkan (Michael Pokorny), Jiaqi Wang (Michael Pokorny), Martin Stehberger (Michael Pokorny), Egor Kretov (Michael Pokorny), Peter Bradshaw (Michael Pokorny), JP Heimonen (Michael Pokorny), Kaustubh Sridhar (Michael Pokorny), Zaki Hossain (Michael Pokorny), Ido Akov (Michael Pokorny), Yury Makarychev (Michael Pokorny), Joanna Tam (Michael Pokorny), Hieu Hoang (Michael Pokorny), David M. Cunningham (Michael Pokorny), Vladimir Goryachev (Michael Pokorny), Demosthenes Patramanis (Michael Pokorny), Michael Krause (Michael Pokorny), Andrew Redenti (Michael Pokorny), David Aldous (Michael Pokorny), Jesyin Lai (Michael Pokorny), Shannon Coleman (Michael Pokorny), Jiangnan Xu (Michael Pokorny), Sangwon Lee (Michael Pokorny), Ilias Magoulas (Michael Pokorny), Sandy Zhao (Michael Pokorny), Ning Tang (Michael Pokorny), Michael K. Cohen (Michael Pokorny), Micah Carroll (Michael Pokorny), Orr Paradise (Michael Pokorny), Jan Hendrik Kirchner (Michael Pokorny), Stefan Steinerberger (Michael Pokorny), Maksym Ovchynnikov (Michael Pokorny), Jason O. Matos (Michael Pokorny), Adithya Shenoy (Michael Pokorny), Michael Wang (Michael Pokorny), Yuzhou Nie (Michael Pokorny), Paolo Giordano (Michael Pokorny), Philipp Petersen (Michael Pokorny), Anna Sztyber-Betley (Michael Pokorny), Paolo Faraboschi (Michael Pokorny), Robin Riblet (Michael Pokorny), Jonathan Crozier (Michael Pokorny), Shiv Halasyamani (Michael Pokorny), Antonella Pinto (Michael Pokorny), Shreyas Verma (Michael Pokorny), Prashant Joshi (Michael Pokorny), Eli Meril (Michael Pokorny), Zheng-Xin Yong (Michael Pokorny), Allison Tee (Michael Pokorny), J\'er\'emy Andr\'eoletti (Michael Pokorny), Orion Weller (Michael Pokorny), Raghav Singhal (Michael Pokorny), Gang Zhang (Michael Pokorny), Alexander Ivanov (Michael Pokorny), Seri Khoury (Michael Pokorny), Nils Gustafsson (Michael Pokorny), Hamid Mostaghimi (Michael Pokorny), Kunvar Thaman (Michael Pokorny), Qijia Chen (Michael Pokorny), Tran Quoc Kh\'anh (Michael Pokorny), Jacob Loader (Michael Pokorny), Stefano Cavalleri (Michael Pokorny), Hannah Szlyk (Michael Pokorny), Zachary Brown (Michael Pokorny), Himanshu Narayan (Michael Pokorny), Jonathan Roberts (Michael Pokorny), William Alley (Michael Pokorny), Kunyang Sun (Michael Pokorny), Ryan Stendall (Michael Pokorny), Max Lamparth (Michael Pokorny), Anka Reuel (Michael Pokorny), Ting Wang (Michael Pokorny), Hanmeng Xu (Michael Pokorny), Pablo Hern\'andez-C\'amara (Michael Pokorny), Freddie Martin (Michael Pokorny), Thomas Preu (Michael Pokorny), Tomek Korbak (Michael Pokorny), Marcus Abramovitch (Michael Pokorny), Dominic Williamson (Michael Pokorny), Ida Bosio (Michael Pokorny), Ziye Chen (Michael Pokorny), Bir\'o B\'alint (Michael Pokorny), Eve J. Y. Lo (Michael Pokorny), Maria In\^es S. Nunes (Michael Pokorny), Yibo Jiang (Michael Pokorny), M Saiful Bari (Michael Pokorny), Peyman Kassani (Michael Pokorny), Zihao Wang (Michael Pokorny), Behzad Ansarinejad (Michael Pokorny), Yewen Sun (Michael Pokorny), Stephane Durand (Michael Pokorny), Guillaume Douville (Michael Pokorny), Daniel Tordera (Michael Pokorny), George Balabanian (Michael Pokorny), Earth Anderson (Michael Pokorny), Lynna Kvistad (Michael Pokorny), Alejandro Jos\'e Moyano (Michael Pokorny), Hsiaoyun Milliron (Michael Pokorny), Ahmad Sakor (Michael Pokorny), Murat Eron (Michael Pokorny), Isaac C. McAlister (Michael Pokorny), Andrew Favre D. O. (Michael Pokorny), Shailesh Shah (Michael Pokorny), Xiaoxiang Zhou (Michael Pokorny), Firuz Kamalov (Michael Pokorny), Ronald Clark (Michael Pokorny), Sherwin Abdoli (Michael Pokorny), Tim Santens (Michael Pokorny), Harrison K Wang (Michael Pokorny), Evan Chen (Michael Pokorny), Alessandro Tomasiello (Michael Pokorny), G. Bruno De Luca (Michael Pokorny), Shi-Zhuo Looi (Michael Pokorny), Vinh-Kha Le (Michael Pokorny), Noam Kolt (Michael Pokorny), Niels M\"undler (Michael Pokorny), Avi Semler (Michael Pokorny), Emma Rodman (Michael Pokorny), Jacob Drori (Michael Pokorny), Carl J Fossum (Michael Pokorny), Luk Gloor (Michael Pokorny), Milind Jagota (Michael Pokorny), Ronak Pradeep (Michael Pokorny), Honglu Fan (Michael Pokorny), Tej Shah (Michael Pokorny), Jonathan Eicher (Michael Pokorny), Michael Chen (Michael Pokorny), Kushal Thaman (Michael Pokorny), William Merrill (Michael Pokorny), Moritz Firsching (Michael Pokorny), Carter Harris (Michael Pokorny), Stefan Ciob\^ac\u{a} (Michael Pokorny), Jason Gross (Michael Pokorny), Rohan Pandey (Michael Pokorny), Ilya Gusev (Michael Pokorny), Adam Jones (Michael Pokorny), Shashank Agnihotri (Michael Pokorny), Pavel Zhelnov (Michael Pokorny), Siranut Usawasutsakorn (Michael Pokorny), Mohammadreza Mofayezi (Michael Pokorny), Alexander Piperski (Michael Pokorny), Marc Carauleanu (Michael Pokorny), David K. Zhang (Michael Pokorny), Kostiantyn Dobarskyi (Michael Pokorny), Dylan Ler (Michael Pokorny), Roman Leventov (Michael Pokorny), Ignat Soroko (Michael Pokorny), Thorben Jansen (Michael Pokorny), Scott Creighton (Michael Pokorny), Pascal Lauer (Michael Pokorny), Joshua Duersch (Michael Pokorny), Vage Taamazyan (Michael Pokorny), Dario Bezzi (Michael Pokorny), Wiktor Morak (Michael Pokorny), Wenjie Ma (Michael Pokorny), William Held (Michael Pokorny), Tran {\DJ}uc Huy (Michael Pokorny), Ruicheng Xian (Michael Pokorny), Armel Randy Zebaze (Michael Pokorny), Mohanad Mohamed (Michael Pokorny), Julian Noah Leser (Michael Pokorny), Michelle X Yuan (Michael Pokorny), Laila Yacar (Michael Pokorny), Johannes Lengler (Michael Pokorny), Katarzyna Olszewska (Michael Pokorny), Hossein Shahrtash (Michael Pokorny), Edson Oliveira (Michael Pokorny), Joseph W. Jackson (Michael Pokorny), Daniel Espinosa Gonzalez (Michael Pokorny), Andy Zou (Michael Pokorny), Muthu Chidambaram (Michael Pokorny), Timothy Manik (Michael Pokorny), Hector Haffenden (Michael Pokorny), Dashiell Stander (Michael Pokorny), Ali Dasouqi (Michael Pokorny), Alexander Shen (Michael Pokorny), Emilien Duc (Michael Pokorny), Bita Golshani (Michael Pokorny), David Stap (Michael Pokorny), Mikalai Uzhou (Michael Pokorny), Alina Borisovna Zhidkovskaya (Michael Pokorny), Lukas Lewark (Michael Pokorny), Miguel Orbegozo Rodriguez (Michael Pokorny), M\'aty\'as Vincze (Michael Pokorny), Dustin Wehr (Michael Pokorny), Colin Tang (Michael Pokorny), Shaun Phillips (Michael Pokorny), Fortuna Samuele (Michael Pokorny), Jiang Muzhen (Michael Pokorny), Fredrik Ekstr\"om (Michael Pokorny), Angela Hammon (Michael Pokorny), Oam Patel (Michael Pokorny), Faraz Farhidi (Michael Pokorny), George Medley (Michael Pokorny), Forough Mohammadzadeh (Michael Pokorny), Madellene Pe\~naflor (Michael Pokorny), Haile Kassahun (Michael Pokorny), Alena Friedrich (Michael Pokorny), Claire Sparrow (Michael Pokorny), Rayner Hernandez Perez (Michael Pokorny), Taom Sakal (Michael Pokorny), Omkar Dhamane (Michael Pokorny), Ali Khajegili Mirabadi (Michael Pokorny), Eric Hallman (Michael Pokorny), Kenchi Okutsu (Michael Pokorny), Mike Battaglia (Michael Pokorny), Mohammad Maghsoudimehrabani (Michael Pokorny), Alon Amit (Michael Pokorny), Dave Hulbert (Michael Pokorny), Roberto Pereira (Michael Pokorny), Simon Weber (Michael Pokorny), Handoko (Michael Pokorny), Anton Peristyy (Michael Pokorny), Stephen Malina (Michael Pokorny), Samuel Albanie (Michael Pokorny), Will Cai (Michael Pokorny), Mustafa Mehkary (Michael Pokorny), Rami Aly (Michael Pokorny), Frank Reidegeld (Michael Pokorny), Anna-Katharina Dick (Michael Pokorny), Cary Friday (Michael Pokorny), Jasdeep Sidhu (Michael Pokorny), Hassan Shapourian (Michael Pokorny), Wanyoung Kim (Michael Pokorny), Mariana Costa (Michael Pokorny), Hubeyb Gurdogan (Michael Pokorny), Brian Weber (Michael Pokorny), Harsh Kumar (Michael Pokorny), Tong Jiang (Michael Pokorny), Arunim Agarwal (Michael Pokorny), Chiara Ceconello (Michael Pokorny), Warren S. Vaz (Michael Pokorny), Chao Zhuang (Michael Pokorny), Haon Park (Michael Pokorny), Andrew R. Tawfeek (Michael Pokorny), Daattavya Aggarwal (Michael Pokorny), Michael Kirchhof (Michael Pokorny), Linjie Dai (Michael Pokorny), Evan Kim (Michael Pokorny), Johan Ferret (Michael Pokorny), Yuzhou Wang (Michael Pokorny), Minghao Yan (Michael Pokorny), Krzysztof Burdzy (Michael Pokorny), Lixin Zhang (Michael Pokorny), Antonio Franca (Michael Pokorny), Diana T. Pham (Michael Pokorny), Kang Yong Loh (Michael Pokorny), Joshua Robinson (Michael Pokorny), Abram Jackson (Michael Pokorny), Shreen Gul (Michael Pokorny), Gunjan Chhablani (Michael Pokorny), Zhehang Du (Michael Pokorny), Adrian Cosma (Michael Pokorny), Jesus Colino (Michael Pokorny), Colin White (Michael Pokorny), Jacob Votava (Michael Pokorny), Vladimir Vinnikov (Michael Pokorny), Ethan Delaney (Michael Pokorny), Petr Spelda (Michael Pokorny), Vit Stritecky (Michael Pokorny), Syed M. Shahid (Michael Pokorny), Jean-Christophe Mourrat (Michael Pokorny), Lavr Vetoshkin (Michael Pokorny), Koen Sponselee (Michael Pokorny), Renas Bacho (Michael Pokorny), Florencia de la Rosa (Michael Pokorny), Xiuyu Li (Michael Pokorny), Guillaume Malod (Michael Pokorny), Leon Lang (Michael Pokorny), Julien Laurendeau (Michael Pokorny), Dmitry Kazakov (Michael Pokorny), Fatimah Adesanya (Michael Pokorny), Julien Portier (Michael Pokorny), Lawrence Hollom (Michael Pokorny), Victor Souza (Michael Pokorny), Yuchen Anna Zhou (Michael Pokorny), Julien Degorre (Michael Pokorny), Yi\u{g}it Yal{\i}n (Michael Pokorny), Gbenga Daniel Obikoya (Michael Pokorny), Luca Arnaboldi (Michael Pokorny), Rai (Michael Pokorny), Filippo Bigi (Quinn), M. C. Bosc\'a (Quinn), Oleg Shumar (Quinn), Kaniuar Bacho (Quinn), Pierre Clavier (Quinn), Gabriel Recchia (Quinn), Mara Popescu (Quinn), Nikita Shulga (Quinn), Ngefor Mildred Tanwie (Quinn), Denis Peskoff (Quinn), Thomas C. H. Lux (Quinn), Ben Rank (Quinn), Colin Ni (Quinn), Matthew Brooks (Quinn), Alesia Yakimchyk (Quinn), Huanxu (Quinn), Liu (Tony), Olle H\"aggstr\"om (Tony), Emil Verkama (Tony), Hans Gundlach (Tony), Leonor Brito-Santana (Tony), Brian Amaro (Tony), Vivek Vajipey (Tony), Rynaa Grover (Tony), Yiyang Fan (Tony), Gabriel Poesia Reis e Silva (Tony), Linwei Xin (Tony), Yosi Kratish (Tony), Jakub {\L}ucki (Tony), Wen-Ding Li (Tony), Sivakanth Gopi (Tony), Andrea Caciolai (Tony), Justin Xu (Tony), Kevin Joseph Scaria (Tony), Freddie Vargus (Tony), Farzad Habibi (Tony), Long (Tony), Lian, Emanuele Rodol\`a, Jules Robins, Vincent Cheng, Tony Fruhauff, Brad Raynor, Hao Qi, Xi Jiang, Ben Segev, Jingxuan Fan, Sarah Martinson, Erik Y. Wang, Kaylie Hausknecht, Michael P. Brenner, Mao Mao, Xinyu Zhang, David Avagian, Eshawn Jessica Scipio, Alon Ragoler, Justin Tan, Blake Sims, Rebeka Plecnik, Aaron Kirtland, Omer Faruk Bodur, D. P. Shinde, Zahra Adoul, Mohamed Zekry, Ali Karakoc, Tania C. B. Santos, Samir Shamseldeen, Loukmane Karim, Anna Liakhovitskaia, Nate Resman, Nicholas Farina, Juan Carlos Gonzalez, Gabe Maayan, Sarah Hoback, Rodrigo De Oliveira Pena, Glen Sherman, Elizabeth Kelley, Hodjat Mariji, Rasoul Pouriamanesh, Wentao Wu, Sandra Mendoza, Ismail Alarab, Joshua Cole, Danyelle Ferreira, Bryan Johnson, Mohammad Safdari, Liangti Dai, Siriphan Arthornthurasuk, Alexey Pronin, Jing Fan, Angel Ramirez-Trinidad, Ashley Cartwright, Daphiny Pottmaier, Omid Taheri, David Outevsky, Stanley Stepanic, Samuel Perry, Luke Askew, Ra\'ul Adri\'an Huerta Rodr\'iguez, Ali M. R. Minissi, Sam Ali, Ricardo Lorena, Krishnamurthy Iyer, Arshad Anil Fasiludeen, Sk Md Salauddin, Murat Islam, Juan Gonzalez, Josh Ducey, Maja Somrak, Vasilios Mavroudis, Eric Vergo, Juehang Qin, Benj\'amin Borb\'as, Eric Chu, Jack Lindsey, Anil Radhakrishnan, Antoine Jallon, I. M. J. McInnis, Pawan Kumar, Laxman Prasad Goswami, Daniel Bugas, Nasser Heydari, Ferenc Jeanplong, Archimedes Apronti, Abdallah Galal, Ng Ze-An, Ankit Singh, Joan of Arc Xavier, Kanu Priya Agarwal, Mohammed Berkani, Benedito Alves de Oliveira Junior, Dmitry Malishev, Nicolas Remy, Taylor D. Hartman, Tim Tarver, Stephen Mensah, Javier Gimenez, Roselynn Grace Montecillo, Russell Campbell, Asankhaya Sharma, Khalida Meer, Xavier Alapont, Deepakkumar Patil, Rajat Maheshwari, Abdelkader Dendane, Priti Shukla, Sergei Bogdanov, S\"oren M\"oller, Muhammad Rehan Siddiqi, Prajvi Saxena, Himanshu Gupta, Innocent Enyekwe, Ragavendran P V, Zienab EL-Wasif, Aleksandr Maksapetyan, Vivien Rossbach, Chris Harjadi, Mohsen Bahaloohoreh, Song Bian, John Lai, Justine Leon Uro, Greg Bateman, Mohamed Sayed, Ahmed Menshawy, Darling Duclosel, Yashaswini Jain, Ashley Aaron, Murat Tiryakioglu, Sheeshram Siddh, Keith Krenek, Alex Hoover, Joseph McGowan, Tejal Patwardhan, Summer Yue, Alexandr Wang, Dan Hendrycks
Abstract: Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai.
URLs: https://lastexam.ai.
Authors: Rahul Atul Bhope, Praveen Venkateswaran, K. R. Jayaram, Vatche Isahagian, Vinod Muthusamy, Nalini Venkatasubramanian
Abstract: Developers using LLMs and LLM-based agents in their applications have provided plenty of anecdotal evidence that in-context-learning (ICL) is fragile. In this paper, we show that in addition to the quantity and quality of examples, the order in which the in-context examples are listed in the prompt affects the output of the LLM and, consequently, their performance. While prior work has explored improving ICL through dataset-dependent techniques, we introduce OptiSeq, a purely inference-time, dataset-free optimization method that efficiently determines the best example order. OptiSeq leverages log probabilities of LLM-generated outputs to systematically prune the search space of possible orderings and recommend the best order(s) by distinguishing orderings that yield high levels of accuracy and those that underperform. Extensive empirical evaluation on multiple LLMs, datasets, and prompts demonstrate that OptiSeq improves accuracy by 5.5 - 10.5 percentage points across multiple tasks.
Authors: Harshith Padigela, Chintan Shah, Dinkar Juyal
Abstract: In this report, we present ML-Dev-Bench, a benchmark aimed at testing agentic capabilities on applied Machine Learning development tasks. While existing benchmarks focus on isolated coding tasks or Kaggle-style competitions, ML-Dev-Bench tests agents' ability to handle the full complexity of ML development workflows. The benchmark assesses performance across critical aspects including dataset handling, model training, improving existing models, debugging, and API integration with popular ML tools. We evaluate three agents - ReAct, Openhands, and AIDE - on a diverse set of 30 tasks, providing insights into their strengths and limitations in handling practical ML development challenges. We open source the benchmark for the benefit of the community at \href{https://github.com/ml-dev-bench/ml-dev-bench}{https://github.com/ml-dev-bench/ml-dev-bench}.
URLs: https://github.com/ml-dev-bench/ml-dev-bench, https://github.com/ml-dev-bench/ml-dev-bench
Authors: Sangjun Han, Taeil Hur, Youngmi Hur, Kathy Sangkyung Lee, Myungyoon Lee, Hyojae Lim
Abstract: The challenge of formal proof generation has a rich history, but with modern techniques, we may finally be at the stage of making actual progress in real-life mathematical problems. This paper explores the integration of ChatGPT and basic searching techniques to simplify generating formal proofs, with a particular focus on the miniF2F dataset. We demonstrate how combining a large language model like ChatGPT with a formal language such as Lean, which has the added advantage of being verifiable, enhances the efficiency and accessibility of formal proof generation. Despite its simplicity, our best-performing Lean-based model surpasses all known benchmarks with a 31.15% pass rate. We extend our experiments to include other datasets and employ alternative language models, showcasing our models' comparable performance in diverse settings and allowing for a more nuanced analysis of our results. Our findings offer insights into AI-assisted formal proof generation, suggesting a promising direction for future research in formal mathematical proof.
Authors: Prakhar Kaushik, Ankit Vaidya, Shravan Chaudhari, Alan Yuille
Abstract: The rapid growth of large models has raised concerns about their environmental impact and equity in accessibility due to significant computational costs. Low-Rank Adapters (LoRA) offer a lightweight solution for finetuning large models, resulting in an abundance of publicly available adapters tailored to diverse domains. We ask: Can these pretrained adapters be leveraged to further streamline adaptation to new tasks while addressing these challenges? We introduce EigenLoRAx, a parameter-efficient finetuning method that recycles existing adapters to create a principal subspace aligned with their shared domain knowledge which can be further augmented with orthogonal basis vectors in low-resource scenarios. This enables rapid adaptation to new tasks by learning only lightweight coefficients on the principal components of the subspace - eliminating the need to finetune entire adapters. EigenLoRAx requires significantly fewer parameters and memory, improving efficiency for both training and inference. Our method demonstrates strong performance across diverse domains and tasks, offering a scalable for edge-based applications, personalization, and equitable deployment of large models in resource-constrained environments.
Authors: Jan Ludziejewski, Maciej Pi\'oro, Jakub Krajewski, Maciej Stefaniak, Micha{\l} Krutul, Jan Ma{\l}a\'snicki, Marek Cygan, Piotr Sankowski, Kamil Adamczewski, Piotr Mi{\l}o\'s, Sebastian Jaszczur
Abstract: Mixture of Experts (MoE) architectures have significantly increased computational efficiency in both research and real-world applications of large-scale machine learning models. However, their scalability and efficiency under memory constraints remain relatively underexplored. In this work, we present joint scaling laws for dense and MoE models, incorporating key factors such as the number of active parameters, dataset size, and the number of experts. Our findings provide a principled framework for selecting the optimal MoE configuration under fixed memory and compute budgets. Surprisingly, we show that MoE models can be more memory-efficient than dense models, contradicting conventional wisdom. To derive and validate the theoretical predictions of our scaling laws, we conduct over 280 experiments with up to 2.7B active parameters and up to 5B total parameters. These results offer actionable insights for designing and deploying MoE models in practical large-scale training scenarios.
Authors: OpenAI, :, Ahmed El-Kishky, Alexander Wei, Andre Saraiva, Borys Minaiev, Daniel Selsam, David Dohan, Francis Song, Hunter Lightman, Ignasi Clavera, Jakub Pachocki, Jerry Tworek, Lorenz Kuhn, Lukasz Kaiser, Mark Chen, Max Schwarzer, Mostafa Rohaninejad, Nat McAleese, o3 contributors, Oleg M\"urk, Rhythm Garg, Rui Shu, Szymon Sidor, Vineet Kosaraju, Wenda Zhou
Abstract: We show that reinforcement learning applied to large language models (LLMs) significantly boosts performance on complex coding and reasoning tasks. Additionally, we compare two general-purpose reasoning models - OpenAI o1 and an early checkpoint of o3 - with a domain-specific system, o1-ioi, which uses hand-engineered inference strategies designed for competing in the 2024 International Olympiad in Informatics (IOI). We competed live at IOI 2024 with o1-ioi and, using hand-crafted test-time strategies, placed in the 49th percentile. Under relaxed competition constraints, o1-ioi achieved a gold medal. However, when evaluating later models such as o3, we find that o3 achieves gold without hand-crafted domain-specific strategies or relaxed constraints. Our findings show that although specialized pipelines such as o1-ioi yield solid improvements, the scaled-up, general-purpose o3 model surpasses those results without relying on hand-crafted inference heuristics. Notably, o3 achieves a gold medal at the 2024 IOI and obtains a Codeforces rating on par with elite human competitors. Overall, these results indicate that scaling general-purpose reinforcement learning, rather than relying on domain-specific techniques, offers a robust path toward state-of-the-art AI in reasoning domains, such as competitive programming.
Authors: Arghadip Das, Arnab Raha, Shamik Kundu, Soumendu Kumar Ghosh, Deepak Mathaikutty, Vijay Raghunathan
Abstract: State-Space Models (SSMs) have emerged as efficient alternatives to transformers for sequential data tasks, offering linear or near-linear scalability with sequence length, making them ideal for long-sequence applications in NLP, vision, and edge AI, including real-time transcription, translation, and contextual search. These applications require lightweight, high-performance models for deployment on resource-constrained devices like laptops and PCs. Designing specialized accelerators for every emerging neural network is costly and impractical; instead, optimizing models for existing NPUs in AI PCs provides a scalable solution. To this end, we propose XAMBA, the first framework to enable and optimize SSMs on commercial off-the-shelf (COTS) state-of-the-art (SOTA) NPUs. XAMBA follows a three-step methodology: (1) enabling SSMs on NPUs, (2) optimizing performance to meet KPI requirements, and (3) trading accuracy for additional performance gains. After enabling SSMs on NPUs, XAMBA mitigates key bottlenecks using CumBA and ReduBA, replacing sequential CumSum and ReduceSum operations with matrix-based computations, significantly improving execution speed and memory efficiency. Additionally, ActiBA enhances performance by approximating expensive activation functions (e.g., Swish, Softplus) using piecewise linear mappings, reducing latency with minimal accuracy loss. Evaluations on an Intel Core Ultra Series 2 AI PC show that XAMBA achieves up to 2.6X speed-up over the baseline. Our implementation is available at https://github.com/arghadippurdue/XAMBA.
Authors: Kehua Chen, Haoyang Shen
Abstract: In this paper, we study semi-supervised Handwritten Mathematical Expression Recognition (HMER) via exploring both labeled data and extra unlabeled data. We propose a novel consistency regularization framework, termed SemiHMER, which introduces dual-branch semi-supervised learning. Specifically, we enforce consistency between the two networks for the same input image. The pseudo-label, generated by one perturbed recognition network, is utilized to supervise the other network using the standard cross-entropy loss. The SemiHMER consistency encourages high similarity between the predictions of the two perturbed networks for the same input image and expands the training data by leveraging unlabeled data with pseudo-labels. We further introduce a weak-to-strong strategy by applying different levels of augmentation to each branch, effectively expanding the training data and enhancing the quality of network training. Additionally, we propose a novel module, the Global Dynamic Counting Module (GDCM), to enhance the performance of the HMER decoder by alleviating recognition inaccuracies in long-distance formula recognition and reducing the occurrence of repeated characters. The experimental results demonstrate that our work achieves significant performance improvements, with an average accuracy increase of 5.47% on CROHME14, 4.87% on CROHME16, and 5.25% on CROHME19, compared to our baselines.
Authors: Yibin Sun, Heitor Murilo Gomes, Bernhard Pfahringer, Albert Bifet
Abstract: The paper explores the challenges of regression analysis in evolving data streams, an area that remains relatively underexplored compared to classification. We propose a standardized evaluation process for regression and prediction interval tasks in streaming contexts. Additionally, we introduce an innovative drift simulation strategy capable of synthesizing various drift types, including the less-studied incremental drift. Comprehensive experiments with state-of-the-art methods, conducted under the proposed process, validate the effectiveness and robustness of our approach.
Authors: Mengchen Fan, Keren Li, Tianyun Zhang, Qing Tian, Baocheng Geng
Abstract: Distributed Learning (DL) enables the training of machine learning models across multiple devices, yet it faces challenges like non-IID data distributions and device capability disparities, which can impede training efficiency. Communication bottlenecks further complicate traditional Federated Learning (FL) setups. To mitigate these issues, we introduce the Personalized Federated Learning with Decentralized Selection Training (PFedDST) framework. PFedDST enhances model training by allowing devices to strategically evaluate and select peers based on a comprehensive communication score. This score integrates loss, task similarity, and selection frequency, ensuring optimal peer connections. This selection strategy is tailored to increase local personalization and promote beneficial peer collaborations to strengthen the stability and efficiency of the training process. Our experiments demonstrate that PFedDST not only enhances model accuracy but also accelerates convergence. This approach outperforms state-of-the-art methods in handling data heterogeneity, delivering both faster and more effective training in diverse and decentralized systems.
Authors: Ao Li, Wei Fang, Hongbo Zhao, Le Lu, Ge Yang, Minfeng Xu
Abstract: In applications of diffusion models, controllable generation is of practical significance, but is also challenging. Current methods for controllable generation primarily focus on modifying the score function of diffusion models, while Mean Reverting (MR) Diffusion directly modifies the structure of the stochastic differential equation (SDE), making the incorporation of image conditions simpler and more natural. However, current training-free fast samplers are not directly applicable to MR Diffusion. And thus MR Diffusion requires hundreds of NFEs (number of function evaluations) to obtain high-quality samples. In this paper, we propose a new algorithm named MRS (MR Sampler) to reduce the sampling NFEs of MR Diffusion. We solve the reverse-time SDE and the probability flow ordinary differential equation (PF-ODE) associated with MR Diffusion, and derive semi-analytical solutions. The solutions consist of an analytical function and an integral parameterized by a neural network. Based on this solution, we can generate high-quality samples in fewer steps. Our approach does not require training and supports all mainstream parameterizations, including noise prediction, data prediction and velocity prediction. Extensive experiments demonstrate that MR Sampler maintains high sampling quality with a speedup of 10 to 20 times across ten different image restoration tasks. Our algorithm accelerates the sampling procedure of MR Diffusion, making it more practical in controllable generation.
Authors: Ziyan Wang, Sizhe Wei, Xiaoming Huo, Hao Wang
Abstract: Diffusion models have made significant advancements in recent years. However, their performance often deteriorates when trained or fine-tuned on imbalanced datasets. This degradation is largely due to the disproportionate representation of majority and minority data in image-text pairs. In this paper, we propose a general fine-tuning approach, dubbed PoGDiff, to address this challenge. Rather than directly minimizing the KL divergence between the predicted and ground-truth distributions, PoGDiff replaces the ground-truth distribution with a Product of Gaussians (PoG), which is constructed by combining the original ground-truth targets with the predicted distribution conditioned on a neighboring text embedding. Experiments on real-world datasets demonstrate that our method effectively addresses the imbalance problem in diffusion models, improving both generation accuracy and quality.
Authors: Mantas Mazeika, Xuwang Yin, Rishub Tamirisa, Jaehyuk Lim, Bruce W. Lee, Richard Ren, Long Phan, Norman Mu, Adam Khoja, Oliver Zhang, Dan Hendrycks
Abstract: As AIs rapidly advance and become more agentic, the risk they pose is governed not only by their capabilities but increasingly by their propensities, including goals and values. Tracking the emergence of goals and values has proven a longstanding problem, and despite much interest over the years it remains unclear whether current AIs have meaningful values. We propose a solution to this problem, leveraging the framework of utility functions to study the internal coherence of AI preferences. Surprisingly, we find that independently-sampled preferences in current LLMs exhibit high degrees of structural coherence, and moreover that this emerges with scale. These findings suggest that value systems emerge in LLMs in a meaningful sense, a finding with broad implications. To study these emergent value systems, we propose utility engineering as a research agenda, comprising both the analysis and control of AI utilities. We uncover problematic and often shocking values in LLM assistants despite existing control measures. These include cases where AIs value themselves over humans and are anti-aligned with specific individuals. To constrain these emergent value systems, we propose methods of utility control. As a case study, we show how aligning utilities with a citizen assembly reduces political biases and generalizes to new scenarios. Whether we like it or not, value systems have already emerged in AIs, and much work remains to fully understand and control these emergent representations.
Authors: Jin Hyun Park, Seyyed Ali Ayati, Yichen Cai
Abstract: The increasing prevalence of microphones in everyday devices and the growing reliance on online services have amplified the risk of acoustic side-channel attacks (ASCAs) targeting keyboards. This study explores deep learning techniques, specifically vision transformers (VTs) and large language models (LLMs), to enhance the effectiveness and applicability of such attacks. We present substantial improvements over prior research, with the CoAtNet model achieving state-of-the-art performance. Our CoAtNet shows a 5.0% improvement for keystrokes recorded via smartphone (Phone) and 5.9% for those recorded via Zoom compared to previous benchmarks. We also evaluate transformer architectures and language models, with the best VT model matching CoAtNet's performance. A key advancement is the introduction of a noise mitigation method for real-world scenarios. By using LLMs for contextual understanding, we detect and correct erroneous keystrokes in noisy environments, enhancing ASCA performance. Additionally, fine-tuned lightweight language models with Low-Rank Adaptation (LoRA) deliver comparable performance to heavyweight models with 67X more parameters. This integration of VTs and LLMs improves the practical applicability of ASCA mitigation, marking the first use of these technologies to address ASCAs and error correction in real-world scenarios.
Authors: Xu Shen, Yixin Liu, Yili Wang, Rui Miao, Yiwei Dai, Shirui Pan, Xin Wang
Abstract: Out-of-distribution (OOD) generalization has emerged as a critical challenge in graph learning, as real-world graph data often exhibit diverse and shifting environments that traditional models fail to generalize across. A promising solution to address this issue is graph invariant learning (GIL), which aims to learn invariant representations by disentangling label-correlated invariant subgraphs from environment-specific subgraphs. However, existing GIL methods face two major challenges: (1) the difficulty of capturing and modeling diverse environments in graph data, and (2) the semantic cliff, where invariant subgraphs from different classes are difficult to distinguish, leading to poor class separability and increased misclassifications. To tackle these challenges, we propose a novel method termed Multi-Prototype Hyperspherical Invariant Learning (MPHIL), which introduces two key innovations: (1) hyperspherical invariant representation extraction, enabling robust and highly discriminative hyperspherical invariant feature extraction, and (2) multi-prototype hyperspherical classification, which employs class prototypes as intermediate variables to eliminate the need for explicit environment modeling in GIL and mitigate the semantic cliff issue. Derived from the theoretical framework of GIL, we introduce two novel objective functions: the invariant prototype matching loss to ensure samples are matched to the correct class prototypes, and the prototype separation loss to increase the distinction between prototypes of different classes in the hyperspherical space. Extensive experiments on 11 OOD generalization benchmark datasets demonstrate that MPHIL achieves state-of-the-art performance, significantly outperforming existing methods across graph data from various domains and with different distribution shifts.
Authors: Zhi Sheng, Yuan Yuan, Yudi Zhang, Depeng Jin, Yong Li
Abstract: Accurate prediction of urban spatiotemporal dynamics is essential for enhancing urban management and decision-making. Existing spatiotemporal prediction models are predominantly deterministic, focusing on primary spatiotemporal patterns. However, those dynamics are highly complex, exhibiting multi-modal distributions that are challenging for deterministic models to capture. In this paper, we highlight the critical role of probabilistic prediction in capturing the uncertainties and complexities inherent in spatiotemporal data. While mainstream probabilistic models can capture uncertainty, they struggle with accurately learning primary patterns and often suffer from computational inefficiency. To address these challenges, we propose CoST, which collaborates deterministic and probabilistic models to improve both predictive accuracy and the ability to handle uncertainty. To achieve this, we design a mean-residual decomposition framework, where the mean value is modeled by a deterministic model, and the residual variations are learned by a probabilistic model, specifically diffusion models. Moreover, we introduce a scale-aware diffusion process, which better accounts for spatially heterogeneous dynamics across different regions. Extensive experiments on eight real-world datasets demonstrate that CoST significantly outperforms existing methods in both deterministic and probabilistic metrics, achieving a 20% improvement with low computational cost. CoST bridges the gap between deterministic precision and probabilistic uncertainty, making a significant advancement in the field of urban spatiotemporal prediction.
Authors: Yuhao Du, Zhuo Li, Pengyu Cheng, Zhihong Chen, Yuejiao Xie, Xiang Wan, Anningzhe Gao
Abstract: Reinforcement Learning from Human Feedback (RLHF) is crucial for aligning Large Language Models (LLMs) with human values. However, RLHF has been continuously challenged by its high complexity in implementation and computation consumption. Even with recent simplifications, such as Direct Preference Optimization (DPO) and Advantage Leftover Lunch (A-LoL), the problems of over-fitting and training instability remain hindering the alignment process from the expected optimal performance. To address the existing challenges, we propose a novel simplification of RLHF from the perspective of variational inference, called $\textbf{V}$ariational $\textbf{A}$lignment with $\textbf{R}$e-weighting ($\textbf{VAR}$). More specifically, by directly minimizing the distribution gap between the learning LLM policy and the optimal solution of RLHF, we transform the alignment objective into a reward-driven re-weighted supervised fine-tuning (SFT) form, which only requires minor adjustment on the SFT loss to obtain noticeable improvement on training stability and effectiveness. On comprehensive alignment and generation benchmarks, our VAR method has numerically achieved competitive performance in LLM alignment helpfulness and harmlessness.
Authors: Zonghao Ying, Deyue Zhang, Zonglei Jing, Yisong Xiao, Quanchen Zou, Aishan Liu, Siyuan Liang, Xiangzheng Zhang, Xianglong Liu, Dacheng Tao
Abstract: Multi-turn jailbreak attacks simulate real-world human interactions by engaging large language models (LLMs) in iterative dialogues, exposing critical safety vulnerabilities. However, existing methods often struggle to balance semantic coherence with attack effectiveness, resulting in either benign semantic drift or ineffective detection evasion. To address this challenge, we propose Reasoning-Augmented Conversation, a novel multi-turn jailbreak framework that reformulates harmful queries into benign reasoning tasks and leverages LLMs' strong reasoning capabilities to compromise safety alignment. Specifically, we introduce an attack state machine framework to systematically model problem translation and iterative reasoning, ensuring coherent query generation across multiple turns. Building on this framework, we design gain-guided exploration, self-play, and rejection feedback modules to preserve attack semantics, enhance effectiveness, and sustain reasoning-driven attack progression. Extensive experiments on multiple LLMs demonstrate that RACE achieves state-of-the-art attack effectiveness in complex conversational scenarios, with attack success rates (ASRs) increasing by up to 96%. Notably, our approach achieves ASRs of 82% and 92% against leading commercial models, OpenAI o1 and DeepSeek R1, underscoring its potency. We release our code at https://github.com/NY1024/RACE to facilitate further research in this critical domain.
Authors: Zongzhao Li, Jiacheng Cen, Bing Su, Wenbing Huang, Tingyang Xu, Yu Rong, Deli Zhao
Abstract: Accurately predicting 3D structures and dynamics of physical systems is crucial in scientific applications. Existing approaches that rely on geometric Graph Neural Networks (GNNs) effectively enforce $\mathrm{E}(3)$-equivariance, but they often fall in leveraging extensive broader information. While direct application of Large Language Models (LLMs) can incorporate external knowledge, they lack the capability for spatial reasoning with guaranteed equivariance. In this paper, we propose EquiLLM, a novel framework for representing 3D physical systems that seamlessly integrates E(3)-equivariance with LLM capabilities. Specifically, EquiLLM comprises four key components: geometry-aware prompting, an equivariant encoder, an LLM, and an equivariant adaptor. Essentially, the LLM guided by the instructive prompt serves as a sophisticated invariant feature processor, while 3D directional information is exclusively handled by the equivariant encoder and adaptor modules. Experimental results demonstrate that EquiLLM delivers significant improvements over previous methods across molecular dynamics simulation, human motion simulation, and antibody design, highlighting its promising generalizability.
Authors: Zihao Wang, Victor Veitch
Abstract: A basic aspiration for interpretability research in large language models is to "localize" semantically meaningful behaviors to particular components within the LLM. There are various heuristics for finding candidate locations within the LLM. Once a candidate localization is found, it can be assessed by editing the internal representations at the corresponding localization and checking whether this induces model behavior that is consistent with the semantic interpretation of the localization. The question we address here is: how strong is the evidence provided by such edits? To evaluate the localization claim, we want to assess the effect of the optimal intervention at a particular location. The key new technical tool is a way of adapting LLM alignment techniques to find such optimal localized edits. With this tool in hand, we give an example where the edit-based evidence for localization appears strong, but where localization clearly fails. Indeed, we find that optimal edits at random localizations can be as effective as aligning the full model. In aggregate, our results suggest that merely observing that localized edits induce targeted changes in behavior provides little to no evidence that these locations actually encode the target behavior.
Authors: Shuqi Liu, Han Wu, Bowei He, Xiongwei Han, Mingxuan Yuan, Linqi Song
Abstract: Recent advances in large language models have led to numerous task-specialized fine-tuned variants, creating a need for efficient model merging techniques that preserve specialized capabilities while avoiding costly retraining. While existing task vector-based merging methods show promise, they typically apply uniform coefficients across all parameters, overlooking varying parameter importance both within and across tasks. We present Sens-Merging, a sensitivity-guided coefficient adjustment method that enhances existing model merging techniques by operating at both task-specific and cross-task levels. Our method analyzes parameter sensitivity within individual tasks and evaluates cross-task transferability to determine optimal merging coefficients. Extensive experiments on Mistral 7B and LLaMA2-7B/13B models demonstrate that Sens-Merging significantly improves performance across general knowledge, mathematical reasoning, and code generation tasks. Notably, when combined with existing merging techniques, our method enables merged models to outperform specialized fine-tuned models, particularly in code generation tasks. Our findings reveal important trade-offs between task-specific and cross-task scalings, providing insights for future model merging strategies.
Authors: Ori Yonay, Tracy Hammond, Tianbao Yang
Abstract: We present Myna, a simple yet effective approach for self-supervised musical representation learning. Built on a contrastive learning framework, Myna introduces two key innovations: (1) the use of a Vision Transformer (ViT) on mel-spectrograms as the backbone and (2) a novel data augmentation strategy, token masking, that masks 90 percent of spectrogram tokens. These innovations deliver both effectiveness and efficiency: (i) Token masking enables a significant increase in per-GPU batch size, from 48 or 120 in prior methods (CLMR, MULE) to 4096. (ii) By avoiding traditional augmentations, Myna retains pitch sensitivity, enhancing performance in tasks like key detection. (iii) The use of vertical patches allows the model to better capture critical features for key detection. Our hybrid model, Myna-22M-Hybrid, processes both 16x16 and 128x2 patches, achieving state-of-the-art results. Trained on a single GPU, it outperforms MULE (62M) on average and rivals MERT-95M, which was trained on 16 and 64 GPUs, respectively. Additionally, it surpasses MERT-95M-public, establishing itself as the best-performing model trained on publicly available data. We release our code and models to promote reproducibility and facilitate future research.
Authors: Sina Montazeri, Haseebullah Jumakhanb, Amir Mirzaeinia
Abstract: This paper investigates the optimization of temporal windows in Financial Deep Reinforcement Learning (DRL) models using 2D Convolutional Neural Networks (CNNs). We introduce a novel approach to treating the temporal field as a hyperparameter and examine its impact on model performance across various datasets and feature arrangements. We introduce a new hyperparameter for the CNN policy, proposing that this temporal field can and should be treated as a hyperparameter for these models. We examine the significance of this temporal field by iteratively expanding the window of observations presented to the CNN policy during the deep reinforcement learning process. Our iterative process involves progressively increasing the observation period from two weeks to twelve weeks, allowing us to examine the effects of different temporal windows on the model's performance. This window expansion is implemented in two settings. In one setting, we rearrange the features in the dataset to group them by company, allowing the model to have a full view of company data in its observation window and CNN kernel. In the second setting, we do not group the features by company, and features are arranged by category. Our study reveals that shorter temporal windows are most effective when no feature rearrangement to group per company is in effect. However, the model will utilize longer temporal windows and yield better performance once we introduce the feature rearrangement. To examine the consistency of our findings, we repeated our experiment on two datasets containing the same thirty companies from the Dow Jones Index but with different features in each dataset and consistently observed the above-mentioned patterns. The result is a trading model significantly outperforming global financial services firms such as the Global X Guru by the established Mirae Asset.
Authors: Kaiyang Wan, Honglin Mu, Rui Hao, Haoran Luo, Tianle Gu, Xiuying Chen
Abstract: Like humans, Large Language Models (LLMs) struggle to generate high-quality long-form text that adheres to strict requirements in a single pass. This challenge is unsurprising, as successful human writing, according to the Cognitive Writing Theory, is a complex cognitive process involving iterative planning, translating, reviewing, and monitoring. Motivated by these cognitive principles, we aim to equip LLMs with human-like cognitive writing capabilities through CogWriter, a novel training-free framework that transforms LLM constrained long-form text generation into a systematic cognitive writing paradigm. Our framework consists of two key modules: (1) a Planning Agent that performs hierarchical planning to decompose the task, and (2) multiple Generation Agents that execute these plans in parallel. The system maintains quality via continuous monitoring and reviewing mechanisms, which evaluate outputs against specified requirements and trigger necessary revisions. CogWriter demonstrates exceptional performance on LongGenBench, a benchmark for complex constrained long-form text generation. Even when using Qwen-2.5-14B as its backbone, CogWriter surpasses GPT-4o by 22% in complex instruction completion accuracy while reliably generating texts exceeding 10,000 words. We hope this cognitive science-inspired approach provides a paradigm for LLM writing advancements: \href{https://github.com/KaiyangWan/CogWriter}{CogWriter}.
Authors: Antonio Purificato, Maria Sofia Bucarelli, Anil Kumar Nelakanti, Andrea Bacciu, Fabrizio Silvestri, Amin Mantrach
Abstract: Reliably labelling data typically requires annotations from multiple human workers. However, humans are far from being perfect. Hence, it is a common practice to aggregate labels gathered from multiple annotators to make a more confident estimate of the true label. Among many aggregation methods, the simple and well known Majority Vote (MV) selects the class label polling the highest number of votes. However, despite its importance, the optimality of MV's label aggregation has not been extensively studied. We address this gap in our work by characterising the conditions under which MV achieves the theoretically optimal lower bound on label estimation error. Our results capture the tolerable limits on annotation noise under which MV can optimally recover labels for a given class distribution. This certificate of optimality provides a more principled approach to model selection for label aggregation as an alternative to otherwise inefficient practices that sometimes include higher experts, gold labels, etc., that are all marred by the same human uncertainty despite huge time and monetary costs. Experiments on both synthetic and real world data corroborate our theoretical findings.
Authors: Ben Liu, Jihan Zhang, Fangquan Lin, Xu Jia, Min Peng
Abstract: Large language models (LLMs) have been increasingly employed in various intelligent educational systems, simulating human tutors to facilitate effective human-machine interaction. However, previous studies often overlook the significance of recognizing and adapting to individual learner characteristics. Such adaptation is crucial for enhancing student engagement and learning efficiency, particularly in mathematics instruction, where diverse learning styles require personalized strategies to promote comprehension and enthusiasm. In this paper, we propose a \textbf{P}erson\textbf{A}lized \textbf{C}onversational tutoring ag\textbf{E}nt (PACE) for mathematics instruction. PACE simulates students' learning styles based on the Felder and Silverman learning style model, aligning with each student's persona. In this way, our PACE can effectively assess the personality of students, allowing to develop individualized teaching strategies that resonate with their unique learning styles. To further enhance students' comprehension, PACE employs the Socratic teaching method to provide instant feedback and encourage deep thinking. By constructing personalized teaching data and training models, PACE demonstrates the ability to identify and adapt to the unique needs of each student, significantly improving the overall learning experience and outcomes. Moreover, we establish multi-aspect evaluation criteria and conduct extensive analysis to assess the performance of personalized teaching. Experimental results demonstrate the superiority of our model in personalizing the educational experience and motivating students compared to existing methods.
Authors: Shengxiang Gao, Jey Han Lau, Jianzhong Qi
Abstract: Knowledge base question answering (KBQA) aims to answer user questions in natural language using rich human knowledge stored in large KBs. As current KBQA methods struggle with unseen knowledge base elements at test time,we introduce SG-KBQA: a novel model that injects schema contexts into entity retrieval and logical form generation to tackle this issue. It uses the richer semantics and awareness of the knowledge base structure provided by schema contexts to enhance generalizability. We show that SG-KBQA achieves strong generalizability, outperforming state-of-the-art models on two commonly used benchmark datasets across a variety of test settings. Our source code is available at https://github.com/gaosx2000/SG_KBQA.
Authors: Rubing Li, Jo\~ao Sedoc, Arun Sundararajan
Abstract: When encountering increasingly frequent performance improvements or cost reductions from a new large language model (LLM), developers of applications leveraging LLMs must decide whether to take advantage of these improvements or stay with older tried-and-tested models. Low perceived switching frictions can lead to choices that do not consider more subtle behavior changes that the transition may induce. Our experiments use a popular game-theoretic behavioral economics model of trust to show stark differences in the trusting behavior of OpenAI's and DeepSeek's models. We highlight a collapse in the economic trust behavior of the o1-mini and o3-mini models as they reconcile profit-maximizing and risk-seeking with future returns from trust, and contrast it with DeepSeek's more sophisticated and profitable trusting behavior that stems from an ability to incorporate deeper concepts like forward planning and theory-of-mind. As LLMs form the basis for high-stakes commercial systems, our results highlight the perils of relying on LLM performance benchmarks that are too narrowly defined and suggest that careful analysis of their hidden fault lines should be part of any organization's AI strategy.
Authors: Ziming Li, Youhuan Li, Yuyu Luo, Guoliang Li, Chuxu Zhang
Abstract: Graph neural networks (GNNs) are powerful deep learning models for graph-structured data, demonstrating remarkable success across diverse domains. Recently, the database (DB) community has increasingly recognized the potentiality of GNNs, prompting a surge of researches focusing on improving database systems through GNN-based approaches. However, despite notable advances, There is a lack of a comprehensive review and understanding of how GNNs could improve DB systems. Therefore, this survey aims to bridge this gap by providing a structured and in-depth overview of GNNs for DB systems. Specifically, we propose a new taxonomy that classifies existing methods into two key categories: (1) Relational Databases, which includes tasks like performance prediction, query optimization, and text-to-SQL, and (2) Graph Databases, addressing challenges like efficient graph query processing and graph similarity computation. We systematically review key methods in each category, highlighting their contributions and practical implications. Finally, we suggest promising avenues for integrating GNNs into Database systems.