Authors: Zhaorun Chen, Zhuokai Zhao, Kai Zhang, Bo Liu, Qi Qi, Yifan Wu, Tarun Kalluri, Sara Cao, Yuanhao Xiong, Haibo Tong, Huaxiu Yao, Hengduo Li, Jiacheng Zhu, Xian Li, Dawn Song, Bo Li, Jason Weston, Dat Huynh
Abstract: While reinforcement learning (RL) can empower large language model (LLM) agents by enabling self-improvement through interaction, its practical adoption remains challenging due to costly rollouts, limited task diversity, unreliable reward signals, and infrastructure complexity, all of which obstruct the collection of scalable experience data. To address these challenges, we introduce DreamGym, the first unified framework designed to synthesize diverse experiences with scalability in mind to enable effective online RL training for autonomous agents. Rather than relying on expensive real-environment rollouts, DreamGym distills environment dynamics into a reasoning-based experience model that derives consistent state transitions and feedback signals through step-by-step reasoning, enabling scalable agent rollout collection for RL. To improve the stability and quality of transitions, DreamGym leverages an experience replay buffer initialized with offline real-world data and continuously enriched with fresh interactions to actively support agent training. To improve knowledge acquisition, DreamGym adaptively generates new tasks that challenge the current agent policy, enabling more effective online curriculum learning. Experiments across diverse environments and agent backbones demonstrate that DreamGym substantially improves RL training, both in fully synthetic settings and in sim-to-real transfer scenarios. On non-RL-ready tasks like WebArena, DreamGym outperforms all baselines by over 30%. And in RL-ready but costly settings, it matches GRPO and PPO performance using only synthetic interactions. When transferring a policy trained purely on synthetic experiences to real-environment RL, DreamGym yields significant additional performance gains while requiring far fewer real-world interactions, providing a scalable warm-start strategy for general-purpose RL.
Authors: Ahmed Mostafa, Raisul Arefin Nahid, Samuel Mulder
Abstract: Tokenization is fundamental in assembly code analysis, impacting intrinsic characteristics like vocabulary size, semantic coverage, and extrinsic performance in downstream tasks. Despite its significance, tokenization in the context of assembly code remains an underexplored area. This study aims to address this gap by evaluating the intrinsic properties of Natural Language Processing (NLP) tokenization models and parameter choices, such as vocabulary size. We explore preprocessing customization options and pre-tokenization rules tailored to the unique characteristics of assembly code. Additionally, we assess their impact on downstream tasks like function signature prediction -- a critical problem in binary code analysis. To this end, we conduct a thorough study on various tokenization models, systematically analyzing their efficiency in encoding assembly instructions and capturing semantic nuances. Through intrinsic evaluations, we compare tokenizers based on tokenization efficiency, vocabulary compression, and representational fidelity for assembly code. Using state-of-the-art pre-trained models such as the decoder-only Large Language Model (LLM) Llama 3.2, the encoder-only transformer BERT, and the encoder-decoder model BART, we evaluate the effectiveness of these tokenizers across multiple performance metrics. Preliminary findings indicate that tokenizer choice significantly influences downstream performance, with intrinsic metrics providing partial but incomplete predictability of extrinsic evaluation outcomes. These results reveal complex trade-offs between intrinsic tokenizer properties and their utility in practical assembly code tasks. Ultimately, this study provides valuable insights into optimizing tokenization models for low-level code analysis, contributing to the robustness and scalability of Natural Language Model (NLM)-based binary analysis workflows.
Authors: Tianning Dong, Luyi Ma, Varun Vasudevan, Jason Cho, Sushant Kumar, Kannan Achan
Abstract: Multimodal Large Language Models (MLLMs) are reshaping how modern agentic systems reason over sequential user-behavior data. However, whether textual or image representations of user behavior data are more effective for maximizing MLLM performance remains underexplored. We present \texttt{BehaviorLens}, a systematic benchmarking framework for assessing modality trade-offs in user-behavior reasoning across six MLLMs by representing transaction data as (1) a text paragraph, (2) a scatter plot, and (3) a flowchart. Using a real-world purchase-sequence dataset, we find that when data is represented as images, MLLMs next-purchase prediction accuracy is improved by 87.5% compared with an equivalent textual representation without any additional computational cost.
Authors: Suraj Prasai, Mengnan Du, Ying Zhang, Fan Yang
Abstract: We develop KnowThyself, an agentic assistant that advances large language model (LLM) interpretability. Existing tools provide useful insights but remain fragmented and code-intensive. KnowThyself consolidates these capabilities into a chat-based interface, where users can upload models, pose natural language questions, and obtain interactive visualizations with guided explanations. At its core, an orchestrator LLM first reformulates user queries, an agent router further directs them to specialized modules, and the outputs are finally contextualized into coherent explanations. This design lowers technical barriers and provides an extensible platform for LLM inspection. By embedding the whole process into a conversational workflow, KnowThyself offers a robust foundation for accessible LLM interpretability.
Authors: Kevin Hong, Kia Karbasi, Gregory Pottie
Abstract: A longstanding goal in computational educational research is to develop explainable knowledge tracing (KT) models. Deep Knowledge Tracing (DKT), which leverages a Recurrent Neural Network (RNN) to predict student knowledge and performance on exercises, has been proposed as a major advancement over traditional KT methods. Several studies suggest that its performance gains stem from its ability to model bidirectional relationships between different knowledge components (KCs) within a course, enabling the inference of a student's understanding of one KC from their performance on others. In this paper, we challenge this prevailing explanation and demonstrate that DKT's strength lies in its implicit ability to model prerequisite relationships as a causal structure, rather than bidirectional relationships. By pruning exercise relation graphs into Directed Acyclic Graphs (DAGs) and training DKT on causal subsets of the Assistments dataset, we show that DKT's predictive capabilities align strongly with these causal structures. Furthermore, we propose an alternative method for extracting exercise relation DAGs using DKT's learned representations and provide empirical evidence supporting our claim. Our findings suggest that DKT's effectiveness is largely driven by its capacity to approximate causal dependencies between KCs rather than simple relational mappings.
Authors: Bram Bult\'e, Ayla Rigouts Terryn
Abstract: Large Language Models (LLMs) are rapidly being adopted by users across the globe, who interact with them in a diverse range of languages. At the same time, there are well-documented imbalances in the training data and optimisation objectives of this technology, raising doubts as to whether LLMs can represent the cultural diversity of their broad user base. In this study, we look at LLMs and cultural values and examine how prompt language and cultural framing influence model responses and their alignment with human values in different countries. We probe 10 LLMs with 63 items from the Hofstede Values Survey Module and World Values Survey, translated into 11 languages, and formulated as prompts with and without different explicit cultural perspectives. Our study confirms that both prompt language and cultural perspective produce variation in LLM outputs, but with an important caveat: While targeted prompting can, to a certain extent, steer LLM responses in the direction of the predominant values of the corresponding countries, it does not overcome the models' systematic bias toward the values associated with a restricted set of countries in our dataset: the Netherlands, Germany, the US, and Japan. All tested models, regardless of their origin, exhibit remarkably similar patterns: They produce fairly neutral responses on most topics, with selective progressive stances on issues such as social tolerance. Alignment with cultural values of human respondents is improved more with an explicit cultural perspective than with a targeted prompt language. Unexpectedly, combining both approaches is no more effective than cultural framing with an English prompt. These findings reveal that LLMs occupy an uncomfortable middle ground: They are responsive enough to changes in prompts to produce variation, but too firmly anchored to specific cultural defaults to adequately represent cultural diversity.
Authors: Zhuowen Yuan, Tao Liu, Yang Yang, Yang Wang, Feng Qi, Kaushik Rangadurai, Bo Li, Shuang Yang
Abstract: Recent LLM-based agents have demonstrated strong capabilities in automated ML engineering. However, they heavily rely on repeated full training runs to evaluate candidate solutions, resulting in significant computational overhead, limited scalability to large search spaces, and slow iteration cycles. To address these challenges, we introduce ArchPilot, a multi-agent system that integrates architecture generation, proxy-based evaluation, and adaptive search into a unified framework. ArchPilot consists of three specialized agents: an orchestration agent that coordinates the search process using a Monte Carlo Tree Search (MCTS)-inspired novel algorithm with a restart mechanism and manages memory of previous candidates; a generation agent that iteratively generates, improves, and debugs candidate architectures; and an evaluation agent that executes proxy training runs, generates and optimizes proxy functions, and aggregates the proxy scores into a fidelity-aware performance metric. This multi-agent collaboration allows ArchPilot to prioritize high-potential candidates with minimal reliance on expensive full training runs, facilitating efficient ML engineering under limited budgets. Experiments on MLE-Bench demonstrate that ArchPilot outperforms SOTA baselines such as AIDE and ML-Master, validating the effectiveness of our multi-agent system.
Authors: Divya Pathak, Harshit Kumar, Anuska Roy, Felix George, Mudit Verma, Pratibha Moogi
Abstract: Multi-Agentic AI systems, powered by large language models (LLMs), are inherently non-deterministic and prone to silent failures such as drift, cycles, and missing details in outputs, which are difficult to detect. We introduce the task of anomaly detection in agentic trajectories to identify these failures and present a dataset curation pipeline that captures user behavior, agent non-determinism, and LLM variation. Using this pipeline, we curate and label two benchmark datasets comprising \textbf{4,275 and 894} trajectories from Multi-Agentic AI systems. Benchmarking anomaly detection methods on these datasets, we show that supervised (XGBoost) and semi-supervised (SVDD) approaches perform comparably, achieving accuracies up to 98% and 96%, respectively. This work provides the first systematic study of anomaly detection in Multi-Agentic AI systems, offering datasets, benchmarks, and insights to guide future research.
Authors: Hirohane Takagi, Gouki Minegishi, Shota Kizawa, Issey Sukeda, Hitomi Yanaka
Abstract: Although behavioral studies have documented numerical reasoning errors in large language models (LLMs), the underlying representational mechanisms remain unclear. We hypothesize that numerical attributes occupy shared latent subspaces and investigate two questions:(1) How do LLMs internally integrate multiple numerical attributes of a single entity? (2)How does irrelevant numerical context perturb these representations and their downstream outputs? To address these questions, we combine linear probing with partial correlation analysis and prompt-based vulnerability tests across models of varying sizes. Our results show that LLMs encode real-world numerical correlations but tend to systematically amplify them. Moreover, irrelevant context induces consistent shifts in magnitude representations, with downstream effects that vary by model size. These findings reveal a vulnerability in LLM decision-making and lay the groundwork for fairer, representation-aware control under multi-attribute entanglement.
Authors: Hao Li, Haotian Chen, Ruoyuan Gong, Juanjuan Wang, Hao Jiang
Abstract: Redistricting plays a central role in shaping how votes are translated into political power. While existing computational methods primarily aim to generate large ensembles of legally valid districting plans, they often neglect the strategic dynamics involved in the selection process. This oversight creates opportunities for partisan actors to cherry-pick maps that, while technically compliant, are politically advantageous. Simply satisfying formal constraints does not ensure fairness when the selection process itself can be manipulated. We propose \textbf{Agentmandering}, a framework that reimagines redistricting as a turn-based negotiation between two agents representing opposing political interests. Drawing inspiration from game-theoretic ideas, particularly the \textit{Choose-and-Freeze} protocol, our method embeds strategic interaction into the redistricting process via large language model (LLM) agents. Agents alternate between selecting and freezing districts from a small set of candidate maps, gradually partitioning the state through constrained and interpretable choices. Evaluation on post-2020 U.S. Census data across all states shows that Agentmandering significantly reduces partisan bias and unfairness, while achieving 2 to 3 orders of magnitude lower variance than standard baselines. These results demonstrate both fairness and stability, especially in swing-state scenarios. Our code is available at https://github.com/Lihaogx/AgentMandering.
Authors: Yuanning Cui, Zequn Sun, Wei Hu, Zhangjie Fu
Abstract: Large language models (LLMs) excel at reasoning but struggle with knowledge-intensive questions due to limited context and parametric knowledge. However, existing methods that rely on finetuned LLMs or GNN retrievers are limited by dataset-specific tuning and scalability on large or unseen graphs. We propose the LLM-KGFR collaborative framework, where an LLM works with a structured retriever, the Knowledge Graph Foundation Retriever (KGFR). KGFR encodes relations using LLM-generated descriptions and initializes entities based on their roles in the question, enabling zero-shot generalization to unseen KGs. To handle large graphs efficiently, it employs Asymmetric Progressive Propagation (APP)- a stepwise expansion that selectively limits high-degree nodes while retaining informative paths. Through node-, edge-, and path-level interfaces, the LLM iteratively requests candidate answers, supporting facts, and reasoning paths, forming a controllable reasoning loop. Experiments demonstrate that LLM-KGFR achieves strong performance while maintaining scalability and generalization, providing a practical solution for KG-augmented reasoning.
Authors: Miguel E. Andres, Vadim Fedorov, Rida Sadek, Enric Spagnolo-Arrizabalaga, Nadescha Trudel
Abstract: Voice AI agents are rapidly transitioning to production deployments, yet systematic methods for ensuring testing reliability remain underdeveloped. Organizations cannot objectively assess whether their testing approaches (internal tools or external platforms) actually work, creating a critical measurement gap as voice AI scales to billions of daily interactions. We present the first systematic framework for evaluating voice AI testing quality through human-centered benchmarking. Our methodology addresses the fundamental dual challenge of testing platforms: generating realistic test conversations (simulation quality) and accurately evaluating agent responses (evaluation quality). The framework combines established psychometric techniques (pairwise comparisons yielding Elo ratings, bootstrap confidence intervals, and permutation tests) with rigorous statistical validation to provide reproducible metrics applicable to any testing approach. To validate the framework and demonstrate its utility, we conducted comprehensive empirical evaluation of three leading commercial platforms focused on Voice AI Testing using 21,600 human judgments across 45 simulations and ground truth validation on 60 conversations. Results reveal statistically significant performance differences with the proposed framework, with the top-performing platform, Evalion, achieving 0.92 evaluation quality measured as f1-score versus 0.73 for others, and 0.61 simulation quality using a league based scoring system (including ties) vs 0.43 for other platforms. This framework enables researchers and organizations to empirically validate the testing capabilities of any platform, providing essential measurement foundations for confident voice AI deployment at scale. Supporting materials are made available to facilitate reproducibility and adoption.
Authors: Claire Yang, Maya Cakmak, Max Kleiman-Weiner
Abstract: Empowerment, a measure of an agent's ability to control its environment, has been proposed as a universal goal-agnostic objective for motivating assistive behavior in AI agents. While multi-human settings like homes and hospitals are promising for AI assistance, prior work on empowerment-based assistance assumes that the agent assists one human in isolation. We introduce an open source multi-human gridworld test suite Disempower-Grid. Using Disempower-Grid, we empirically show that assistive RL agents optimizing for one human's empowerment can significantly reduce another human's environmental influence and rewards - a phenomenon we formalize as disempowerment. We characterize when disempowerment occurs in these environments and show that joint empowerment mitigates disempowerment at the cost of the user's reward. Our work reveals a broader challenge for the AI alignment community: goal-agnostic objectives that seem aligned in single-agent settings can become misaligned in multi-agent contexts.
Authors: Alan Seroul, Th\'eo Fagnoni, In\`es Adnani, Dana O. Mohamed, Phillip Kingston
Abstract: This paper introduces the Opus Workflow Evaluation Framework, a probabilistic-normative formulation for quantifying Workflow quality and efficiency. It integrates notions of correctness, reliability, and cost into a coherent mathematical model that enables direct comparison, scoring, and optimization of Workflows. The framework combines the Opus Workflow Reward, a probabilistic function estimating expected performance through success likelihood, resource usage, and output gain, with the Opus Workflow Normative Penalties, a set of measurable functions capturing structural and informational quality across Cohesion, Coupling, Observability, and Information Hygiene. It supports automated Workflow assessment, ranking, and optimization within modern automation systems such as Opus and can be integrated into Reinforcement Learning loops to guide Workflow discovery and refinement. In this paper, we introduce the Opus Workflow Reward model that formalizes Workflow success as a probabilistic expectation over costs and outcomes. We define measurable Opus Workflow Normative Penalties capturing structural, semantic, and signal-related properties of Workflows. Finally, we propose a unified optimization formulation for identifying and ranking optimal Workflows under joint Reward-Penalty trade-offs.
Authors: Zhengru Fang, Yu Guo, Jingjing Wang, Yuang Zhang, Haonan An, Yinhai Wang, Yuguang Fang
Abstract: Sharing and reconstructing a consistent spatial memory is a critical challenge in multi-agent systems, where partial observability and limited bandwidth often lead to catastrophic failures in coordination. We introduce a multi-agent predictive coding framework that formulate coordination as the minimization of mutual uncertainty among agents. Instantiated as an information bottleneck objective, it prompts agents to learn not only who and what to communicate but also when. At the foundation of this framework lies a grid-cell-like metric as internal spatial coding for self-localization, emerging spontaneously from self-supervised motion prediction. Building upon this internal spatial code, agents gradually develop a bandwidth-efficient communication mechanism and specialized neural populations that encode partners' locations: an artificial analogue of hippocampal social place cells (SPCs). These social representations are further enacted by a hierarchical reinforcement learning policy that actively explores to reduce joint uncertainty. On the Memory-Maze benchmark, our approach shows exceptional resilience to bandwidth constraints: success degrades gracefully from 73.5% to 64.4% as bandwidth shrinks from 128 to 4 bits/step, whereas a full-broadcast baseline collapses from 67.6% to 28.6%. Our findings establish a theoretically principled and biologically plausible basis for how complex social representations emerge from a unified predictive drive, leading to social collective intelligence.
Authors: Zeng Zhiyuan, Jiashuo Liu, Zhangyue Yin, Ge Zhang, Wenhao Huang, Xipeng Qiu
Abstract: While Reinforcement Learning for Verifiable Rewards (RLVR) is powerful for training large reasoning models, its training dynamics harbor a critical challenge: RL overfitting, where models gain training rewards but lose generalization. Our analysis reveals this is driven by policy over-specialization and catastrophic forgetting of diverse solutions generated during training. Standard optimization discards this valuable inter-step policy diversity. To address this, we introduce RLoop, a self-improving framework built on iterative policy initialization. RLoop transforms the standard training process into a virtuous cycle: it first uses RL to explore the solution space from a given policy, then filters the successful trajectories to create an expert dataset. This dataset is used via Rejection-sampling Fine-Tuning (RFT) to refine the initial policy, creating a superior starting point for the next iteration. This loop of exploration and exploitation via iterative re-initialization effectively converts transient policy variations into robust performance gains. Our experiments show RLoop mitigates forgetting and substantially improves generalization, boosting average accuracy by 9% and pass@32 by over 15% compared to vanilla RL.
Authors: Jian Mu, Chaoyun Zhang, Chiming Ni, Lu Wang, Bo Qiao, Kartik Mathur, Qianhui Wu, Yuhang Xie, Xiaojun Ma, Mengyu Zhou, Si Qin, Liqun Li, Yu Kang, Minghua Ma, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang
Abstract: We introduce GUI-360$^\circ$, a large-scale, comprehensive dataset and benchmark suite designed to advance computer-using agents (CUAs). CUAs present unique challenges and is constrained by three persistent gaps: a scarcity of real-world CUA tasks, the lack of automated collection-and-annotation pipelines for multi-modal trajectories, and the absence of a unified benchmark that jointly evaluates GUI grounding, screen parsing, and action prediction. GUI-360$^\circ$ addresses these gaps with an LLM-augmented, largely automated pipeline for query sourcing, environment-template construction, task instantiation, batched execution, and LLM-driven quality filtering. The released corpus contains over 1.2M executed action steps across thousands of trajectories in popular Windows office applications, and includes full-resolution screenshots, accessibility metadata when available, instantiated goals, intermediate reasoning traces, and both successful and failed action trajectories. The dataset supports three canonical tasks, GUI grounding, screen parsing, and action prediction, and a hybrid GUI+API action space that reflects modern agent designs. Benchmarking state-of-the-art vision--language models on GUI-360$^\circ$ reveals substantial out-of-the-box shortcomings in grounding and action prediction; supervised fine-tuning and reinforcement learning yield significant gains but do not close the gap to human-level reliability. We release GUI-360$^\circ$ and accompanying code to facilitate reproducible research and accelerate progress on robust desktop CUAs. The full dataset has been made public on https://huggingface.co/datasets/vyokky/GUI-360.
Authors: Jacob Lysn{\ae}s-Larsen, Marte Eggen, Inga Str\"umke
Abstract: In explainable AI, Concept Activation Vectors (CAVs) are typically obtained by training linear classifier probes to detect human-understandable concepts as directions in the activation space of deep neural networks. It is widely assumed that a high probe accuracy indicates a CAV faithfully representing its target concept. However, we show that the probe's classification accuracy alone is an unreliable measure of concept alignment, i.e., the degree to which a CAV captures the intended concept. In fact, we argue that probes are more likely to capture spurious correlations than they are to represent only the intended concept. As part of our analysis, we demonstrate that deliberately misaligned probes constructed to exploit spurious correlations, achieve an accuracy close to that of standard probes. To address this severe problem, we introduce a novel concept localization method based on spatial linear attribution, and provide a comprehensive comparison of it to existing feature visualization techniques for detecting and mitigating concept misalignment. We further propose three classes of metrics for quantitatively assessing concept alignment: hard accuracy, segmentation scores, and augmentation robustness. Our analysis shows that probes with translation invariance and spatial alignment consistently increase concept alignment. These findings highlight the need for alignment-based evaluation metrics rather than probe accuracy, and the importance of tailoring probes to both the model architecture and the nature of the target concept.
Authors: Tim Beyer, Jonas Dornbusch, Jakob Steimle, Moritz Ladenburger, Leo Schwinn, Stephan G\"unnemann
Abstract: The rapid expansion of research on Large Language Model (LLM) safety and robustness has produced a fragmented and oftentimes buggy ecosystem of implementations, datasets, and evaluation methods. This fragmentation makes reproducibility and comparability across studies challenging, hindering meaningful progress. To address these issues, we introduce AdversariaLLM, a toolbox for conducting LLM jailbreak robustness research. Its design centers on reproducibility, correctness, and extensibility. The framework implements twelve adversarial attack algorithms, integrates seven benchmark datasets spanning harmfulness, over-refusal, and utility evaluation, and provides access to a wide range of open-weight LLMs via Hugging Face. The implementation includes advanced features for comparability and reproducibility such as compute-resource tracking, deterministic results, and distributional evaluation techniques. \name also integrates judging through the companion package JudgeZoo, which can also be used independently. Together, these components aim to establish a robust foundation for transparent, comparable, and reproducible research in LLM safety.
Authors: Jiahao Zhao, Luxin Xu, Minghuan Tan, Lichao Zhang, Ahmadreza Argha, Hamid Alinejad-Rokny, Min Yang
Abstract: Numerous medical systems powered by Large Language Models (LLMs) have achieved remarkable progress in diverse healthcare tasks. However, research on their medication safety remains limited due to the lack of real world datasets, constrained by privacy and accessibility issues. Moreover, evaluation of LLMs in realistic clinical consultation settings, particularly regarding medication safety, is still underexplored. To address these gaps, we propose a framework that simulates and evaluates clinical consultations to systematically assess the medication safety capabilities of LLMs. Within this framework, we generate inquiry diagnosis dialogues with embedded medication risks and construct a dedicated medication safety database, RxRisk DB, containing 6,725 contraindications, 28,781 drug interactions, and 14,906 indication-drug pairs. A two-stage filtering strategy ensures clinical realism and professional quality, resulting in the benchmark RxSafeBench with 2,443 high-quality consultation scenarios. We evaluate leading open-source and proprietary LLMs using structured multiple choice questions that test their ability to recommend safe medications under simulated patient contexts. Results show that current LLMs struggle to integrate contraindication and interaction knowledge, especially when risks are implied rather than explicit. Our findings highlight key challenges in ensuring medication safety in LLM-based systems and provide insights into improving reliability through better prompting and task-specific tuning. RxSafeBench offers the first comprehensive benchmark for evaluating medication safety in LLMs, advancing safer and more trustworthy AI-driven clinical decision support.
Authors: Nick Oh, Fernand Gobet
Abstract: Test-time reasoning architectures such as those following the Generate-Verify paradigm -- where a model iteratively refines or verifies its own generated outputs -- prioritise generation and verification but exclude the monitoring processes that determine when and how reasoning should begin. This omission may contribute to the prefix dominance trap, in which models commit early to suboptimal reasoning paths and seldom recover, yielding roughly 20% accuracy loss. We address this architectural gap by formalising Flavell's and Nelson and Narens' metacognitive theories into computational specifications, proposing the Monitor-Generate-Verify (MGV) framework. MGV extends the Generate-Verify paradigm by adding explicit monitoring that captures metacognitive experiences (from difficulty assessments to confidence judgements) before generation begins and refines future monitoring through verification feedback. Though we present no empirical validation, this work provides the first systematic computational translation of foundational metacognitive theories, offering a principled vocabulary for understanding reasoning system failures and suggesting specific architectural interventions for future test-time reasoning designs.
Authors: Chanwoo Park, Ziyang Chen, Asuman Ozdaglar, Kaiqing Zhang
Abstract: Large language models (LLMs) are increasingly deployed as "agents" for decision-making (DM) in interactive and dynamic environments. Yet, since they were not originally designed for DM, recent studies show that LLMs can struggle even in basic online DM problems, failing to achieve low regret or an effective exploration-exploitation tradeoff. To address this, we introduce Iterative Regret-Minimization Fine-Tuning (Iterative RMFT), a post-training procedure that repeatedly distills low-regret decision trajectories back into the base model. At each iteration, the model rolls out multiple decision trajectories, selects the k-lowest regret ones, and fine-tunes itself on them. Unlike prior methods that (a) distill action sequences from known DM algorithms or (b) rely on manually crafted chain-of-thought templates, our approach leverages the regret metric to elicit the model's own DM ability and reasoning rationales. This reliance on model-generated reasoning avoids rigid output engineering and provides more flexible, natural-language training signals. Empirical results show that Iterative RMFT improves LLMs' DM performance across diverse models - from Transformers with numerical input/output, to open-weight LLMs, and advanced closed-weight models like GPT-4o mini. Its flexibility in output and reasoning formats enables generalization across tasks with varying horizons, action spaces, reward processes, and natural-language contexts. Finally, we provide theoretical insight showing that a single-layer Transformer under this paradigm can act as a no-regret learner in a simplified setting. Overall, Iterative RMFT offers a principled and general post-training framework for enhancing LLMs' decision-making capabilities.
Authors: Anisha Garg, Ganesh Venkatesh
Abstract: Group-relative Policy Optimization's (GRPO) simplicity makes it highly desirable for adapting LLMs to become experts at specific tasks. But this simplicity also makes it ill-specified as we seek to enhance RL training with richer, non-binary feedback. When using ordinal rewards to give partial credit, GRPO's simplicity starts to hurt, as its group-average baseline often assigns a positive advantage to failed trajectories and reinforces incorrect behavior. We introduce Correctness Relative Policy Optimization (CoRPO), a new formulation that solves this flaw. CoRPO uses an adaptive baseline that enforces a minimum quality threshold, ensuring failed solutions are never positively reinforced. Once the policy consistently meets this threshold, the baseline automatically transitions to a relative preference mode, pushing the model to find optimal solutions rather than just "acceptable" ones. We empirically validate CoRPO on a code verification task, where it demonstrates more stable convergence and better out-of-domain generalization. This work represents a critical step in our broader research program to enable LLMs to learn genuinely new capabilities through reinforcement learning. We achieve this by enabling LLMs to learn from rich, multi-dimensional feedback - progressing from binary to ordinal rewards in this work, and onward to denser, per-step supervision.
Authors: Carnot Braun, Rafael O. Jarczewski, Gabriel U. Talasso, Leandro A. Villas, Allan M. de Souza
Abstract: Traditional vehicle routing systems efficiently optimize singular metrics like time or distance, and when considering multiple metrics, they need more processes to optimize . However, they lack the capability to interpret and integrate the complex, semantic, and dynamic contexts of human drivers, such as multi-step tasks, situational constraints, or urgent needs. This paper introduces and evaluates PAVe (Personalized Agentic Vehicular Routing), a hybrid agentic assistant designed to augment classical pathfinding algorithms with contextual reasoning. Our approach employs a Large Language Model (LLM) agent that operates on a candidate set of routes generated by a multi-objective (time, CO2) Dijkstra algorithm. The agent evaluates these options against user-provided tasks, preferences, and avoidance rules by leveraging a pre-processed geospatial cache of urban Points of Interest (POIs). In a benchmark of realistic urban scenarios, PAVe successfully used complex user intent into appropriate route modifications, achieving over 88% accuracy in its initial route selections with a local model. We conclude that combining classical routing algorithms with an LLM-based semantic reasoning layer is a robust and effective approach for creating personalized, adaptive, and scalable solutions for urban mobility optimization.
Authors: Lars Krupp, Daniel Gei{\ss}ler, Vishal Banwari, Paul Lukowicz, Jakob Karolus
Abstract: Web agents, like OpenAI's Operator and Google's Project Mariner, are powerful agentic systems pushing the boundaries of Large Language Models (LLM). They can autonomously interact with the internet at the user's behest, such as navigating websites, filling search masks, and comparing price lists. Though web agent research is thriving, induced sustainability issues remain largely unexplored. To highlight the urgency of this issue, we provide an initial exploration of the energy and $CO_2$ cost associated with web agents from both a theoretical -via estimation- and an empirical perspective -by benchmarking. Our results show how different philosophies in web agent creation can severely impact the associated expended energy, and that more energy consumed does not necessarily equate to better results. We highlight a lack of transparency regarding disclosing model parameters and processes used for some web agents as a limiting factor when estimating energy consumption. Our work contributes towards a change in thinking of how we evaluate web agents, advocating for dedicated metrics measuring energy consumption in benchmarks.
Authors: Andrea Cera Palatsi, Samuel Martin-Gutierrez, Ana S. Cardenal, Max Pellert
Abstract: Large language models (LLMs) are increasingly used both to make decisions in domains such as health, education and law, and to simulate human behavior. Yet how closely LLMs mirror actual human decision-making remains poorly understood. This gap is critical: misalignment could produce harmful outcomes in practical applications, while failure to replicate human behavior renders LLMs ineffective for social simulations. Here, we address this gap by developing a digital twin of game-theoretic experiments and introducing a systematic prompting and probing framework for machine-behavioral evaluation. Testing three open-source models (Llama, Mistral and Qwen), we find that Llama reproduces human cooperation patterns with high fidelity, capturing human deviations from rational choice theory, while Qwen aligns closely with Nash equilibrium predictions. Notably, we achieved population-level behavioral replication without persona-based prompting, simplifying the simulation process. Extending beyond the original human-tested games, we generate and preregister testable hypotheses for novel game configurations outside the original parameter grid. Our findings demonstrate that appropriately calibrated LLMs can replicate aggregate human behavioral patterns and enable systematic exploration of unexplored experimental spaces, offering a complementary approach to traditional research in the social and behavioral sciences that generates new empirical predictions about human social decision-making.
Authors: Zihang Ding, Kun Zhang
Abstract: Urban surface water flooding, triggered by intense rainfall overwhelming drainage systems, is increasingly frequent and widespread. While flood prediction and monitoring in high spatial-temporal resolution are desired, practical constraints in time, budget, and technology hinder its full implementation. How to monitor urban drainage networks and predict flow conditions under constrained resource is a major challenge. This study presents a data-driven sparse sensing (DSS) framework, integrated with EPA-SWMM, to optimize sensor placement and reconstruct peak flowrates in a stormwater system, using the Woodland Avenue catchment in Duluth, Minnesota, as a case study. We utilized a SWMM model to generate a training dataset of peak flowrate profiles across the stormwater network. Furthermore, we applied DSS - leveraging singular value decomposition for dimensionality reduction and QR factorization for sensor allocation - to identify the optimal monitoring nodes based on the simulated training dataset. We then validated the representativeness of these identified monitoring nodes by comparing the DSS-reconstructed peak flowrate profiles with those obtained from SWMM. Three optimally placed sensors among 77 nodes achieved satisfactory reconstruction performance with Nash-Sutcliffe Efficiency (NSE) values of 0.92-0.95 (25th to 75th percentiles). In addition, the model showed good robustness to uncertainty in measurements. Its robustness to sensor failures is location-dependent and improves with the number of sensors deployed. The framework balances computational efficiency and physical interpretability, enabling high-accuracy flow reconstruction with minimal sensors. This DSS framework can be further integrated with predictive models to realize flood early warning and real-time control under limited sensing and monitoring resource.
Authors: Atsuyuki Miyai, Mashiro Toyooka, Takashi Otonari, Zaiying Zhao, Kiyoharu Aizawa
Abstract: Understanding the current capabilities and risks of AI Scientist systems is essential for ensuring trustworthy and sustainable AI-driven scientific progress while preserving the integrity of the academic ecosystem. To this end, we develop Jr. AI Scientist, a state-of-the-art autonomous AI scientist system that mimics the core research workflow of a novice student researcher: Given the baseline paper from the human mentor, it analyzes its limitations, formulates novel hypotheses for improvement, validates them through rigorous experimentation, and writes a paper with the results. Unlike previous approaches that assume full automation or operate on small-scale code, Jr. AI Scientist follows a well-defined research workflow and leverages modern coding agents to handle complex, multi-file implementations, leading to scientifically valuable contributions. For evaluation, we conducted automated assessments using AI Reviewers, author-led evaluations, and submissions to Agents4Science, a venue dedicated to AI-driven scientific contributions. The findings demonstrate that Jr. AI Scientist generates papers receiving higher review scores than existing fully automated systems. Nevertheless, we identify important limitations from both the author evaluation and the Agents4Science reviews, indicating the potential risks of directly applying current AI Scientist systems and key challenges for future research. Finally, we comprehensively report various risks identified during development. We hope these insights will deepen understanding of current progress and risks in AI Scientist development.
Authors: Daniel Gomm, Cornelius Wolff, Madelon Hulsebos
Abstract: Natural language interfaces to tabular data must handle ambiguities inherent to queries. Instead of treating ambiguity as a deficiency, we reframe it as a feature of cooperative interaction, where the responsibility of query specification is shared among the user and the system. We develop a principled framework distinguishing cooperative queries, i.e., queries that yield a resolvable interpretation, from uncooperative queries that cannot be resolved. Applying the framework to evaluations for tabular question answering and analysis, we analyze the queries in 15 popular datasets, and observe an uncontrolled mixing of query types neither adequate for evaluating a system's execution accuracy nor for evaluating interpretation capabilities. Our framework and analysis of queries shifts the perspective from fixing ambiguity to embracing cooperation in resolving queries. This reflection enables more informed design and evaluation for natural language interfaces for tabular data, for which we outline implications and directions for future research.
Authors: Soham De, Lodewijk Gelauff, Ashish Goel, Smitha Milli, Ariel Procaccia, Alice Siu
Abstract: A central feature of many deliberative processes, such as citizens' assemblies and deliberative polls, is the opportunity for participants to engage directly with experts. While participants are typically invited to propose questions for expert panels, only a limited number can be selected due to time constraints. This raises the challenge of how to choose a small set of questions that best represent the interests of all participants. We introduce an auditing framework for measuring the level of representation provided by a slate of questions, based on the social choice concept known as justified representation (JR). We present the first algorithms for auditing JR in the general utility setting, with our most efficient algorithm achieving a runtime of $O(mn\log n)$, where $n$ is the number of participants and $m$ is the number of proposed questions. We apply our auditing methods to historical deliberations, comparing the representativeness of (a) the actual questions posed to the expert panel (chosen by a moderator), (b) participants' questions chosen via integer linear programming, (c) summary questions generated by large language models (LLMs). Our results highlight both the promise and current limitations of LLMs in supporting deliberative processes. By integrating our methods into an online deliberation platform that has been used for over hundreds of deliberations across more than 50 countries, we make it easy for practitioners to audit and improve representation in future deliberations.
Authors: Narjes Nourzad, Hanqing Yang, Shiyu Chen, Carlee Joe-Wong
Abstract: Cooperative multi-agent planning requires agents to make joint decisions with partial information and limited communication. Coordination at the trajectory level often fails, as small deviations in timing or movement cascade into conflicts. Symbolic planning mitigates this challenge by raising the level of abstraction and providing a minimal vocabulary of actions that enable synchronization and collective progress. We present DR. WELL, a decentralized neurosymbolic framework for cooperative multi-agent planning. Cooperation unfolds through a two-phase negotiation protocol: agents first propose candidate roles with reasoning and then commit to a joint allocation under consensus and environment constraints. After commitment, each agent independently generates and executes a symbolic plan for its role without revealing detailed trajectories. Plans are grounded in execution outcomes via a shared world model that encodes the current state and is updated as agents act. By reasoning over symbolic plans rather than raw trajectories, DR. WELL avoids brittle step-level alignment and enables higher-level operations that are reusable, synchronizable, and interpretable. Experiments on cooperative block-push tasks show that agents adapt across episodes, with the dynamic world model capturing reusable patterns and improving task completion rates and efficiency. Experiments on cooperative block-push tasks show that our dynamic world model improves task completion and efficiency through negotiation and self-refinement, trading a time overhead for evolving, more efficient collaboration strategies.
Authors: Yu Feng, Nathaniel Weir, Kaj Bostrom, Sam Bayless, Darion Cassel, Sapana Chaudhary, Benjamin Kiesl-Reiter, Huzefa Rangwala
Abstract: LLMs can perform multi-step reasoning through Chain-of-Thought (CoT), but they cannot reliably verify their own logic. Even when they reach correct answers, the underlying reasoning may be flawed, undermining trust in high-stakes scenarios. To mitigate this issue, we introduce VeriCoT, a neuro-symbolic method that extracts and verifies formal logical arguments from CoT reasoning. VeriCoT formalizes each CoT reasoning step into first-order logic and identifies premises that ground the argument in source context, commonsense knowledge, or prior reasoning steps. The symbolic representation enables automated solvers to verify logical validity while the NL premises allow humans and systems to identify ungrounded or fallacious reasoning steps. Experiments on the ProofWriter, LegalBench, and BioASQ datasets show VeriCoT effectively identifies flawed reasoning, and serves as a strong predictor of final answer correctness. We also leverage VeriCoT's verification signal for (1) inference-time self-reflection, (2) supervised fine-tuning (SFT) on VeriCoT-distilled datasets and (3) preference fine-tuning (PFT) with direct preference optimization (DPO) using verification-based pairwise rewards, further improving reasoning validity and accuracy.
Authors: Fatimo Adenike Adeniya (York St John University, London Campus, London, United Kingdom)
Abstract: Cyberattacks on e-commerce platforms have grown in sophistication, threatening consumer trust and operational continuity. This research presents a hybrid analytical framework that integrates statistical modelling and machine learning for detecting and forecasting cyberattack patterns in the e-commerce domain. Using the Verizon Community Data Breach (VCDB) dataset, the study applies Auto ARIMA for temporal forecasting and significance testing, including a Mann-Whitney U test (U = 2579981.5, p = 0.0121), which confirmed that holiday shopping events experienced significantly more severe cyberattacks than non-holiday periods. ANOVA was also used to examine seasonal variation in threat severity, while ensemble machine learning models (XGBoost, LightGBM, and CatBoost) were employed for predictive classification. Results reveal recurrent attack spikes during high-risk periods such as Black Friday and holiday seasons, with breaches involving Personally Identifiable Information (PII) exhibiting elevated threat indicators. Among the models, CatBoost achieved the highest performance (accuracy = 85.29%, F1 score = 0.2254, ROC AUC = 0.8247). The framework uniquely combines seasonal forecasting with interpretable ensemble learning, enabling temporal risk anticipation and breach-type classification. Ethical considerations, including responsible use of sensitive data and bias assessment, were incorporated. Despite class imbalance and reliance on historical data, the study provides insights for proactive cybersecurity resource allocation and outlines directions for future real-time threat detection research.
Authors: Atena Khoshkonesh, Mohsen Mohammadagha, Navid Ebrahimi
Abstract: Persistent cost and schedule deviations remain a major challenge in the U.S. construction industry, revealing the limitations of deterministic CPM and static document-based estimating. This study presents an integrated 4D/5D digital-twin framework that couples Building Information Modeling (BIM) with natural-language processing (NLP)-based cost mapping, computer-vision (CV)-driven progress measurement, Bayesian probabilistic CPM updating, and deep-reinforcement-learning (DRL) resource-leveling. A nine-month case implementation on a Dallas-Fort Worth mid-rise project demonstrated measurable gains in accuracy and efficiency: 43% reduction in estimating labor, 6% reduction in overtime, and 30% project-buffer utilization, while maintaining an on-time finish at 128 days within P50-P80 confidence bounds. The digital-twin sandbox also enabled real-time "what-if" forecasting and traceable cost-schedule alignment through a 5D knowledge graph. Findings confirm that integrating AI-based analytics with probabilistic CPM and DRL enhances forecasting precision, transparency, and control resilience. The validated workflow establishes a practical pathway toward predictive, adaptive, and auditable construction management.
Authors: Chenyu Hou, Hua Yu, Gaoxia Zhu, John Derek Anas, Jiao Liu, Yew Soon Ong
Abstract: Computational Thinking (CT) is a foundational problem-solving skill, and gamified programming environments are a widely adopted approach to cultivating it. While large language models (LLMs) provide on-demand programming support, current applications rarely foster CT development. We present MazeMate, an LLM-powered chatbot embedded in a 3D Maze programming game, designed to deliver adaptive, context-sensitive scaffolds aligned with CT processes in maze solving and maze design. We report on the first classroom implementation with 247 undergraduates. Students rated MazeMate as moderately helpful, with higher perceived usefulness for maze solving than for maze design. Thematic analysis confirmed support for CT processes such as decomposition, abstraction, and algorithmic thinking, while also revealing limitations in supporting maze design, including mismatched suggestions and fabricated algorithmic solutions. These findings demonstrate the potential of LLM-based scaffolding to support CT and underscore directions for design refinement to enhance MazeMate usability in authentic classrooms.
Authors: Sanidhya Vijayvargiya, Rahul Lokesh
Abstract: On-device AI agents offer the potential for personalized, low-latency assistance, but their deployment is fundamentally constrained by limited memory capacity, which restricts usable context. This reduced practical context window creates a trade-off between supporting rich, stateful interactions with complex tool capabilities and maintaining on-device feasibility. We break this trade-off with a framework for context-efficient on-device agents, driven by three synergistic optimizations (1) a dynamic memory system using specialized LoRA adapters to distill conversational history into a compressed, and structured Context State Object; (2) a minimalist serialization format for tool schemas to minimize token overhead per tool; and (3) a just-in-time schema-passing mechanism that loads full tool definitions only upon tool selection. We instantiate this framework by adapting a 3B parameter SLM to context-efficient trajectories and rigorously evaluate it against a conventional baseline on complex user tasks. Our agent matches, or exceeds, the performance of a conventional baseline while dramatically compressing context, achieving more than a 6-fold reduction in initial system prompt context and a 10- to 25-fold reduction in context growth rate based on the interaction verbosity, demonstrating that strategic context management is key to unlocking capable and persistent on-device AI.
Authors: Zhiyin Zhou
Abstract: Large language models are moving beyond transactional question answering to act as companions, coaches, mediators, and curators that scaffold human growth, decision-making, and well-being. This paper proposes a role-based framework for human-centered LLM support systems, compares real deployments across domains, and identifies cross-cutting design principles: transparency, personalization, guardrails, memory with privacy, and a balance of empathy and reliability. It outlines evaluation metrics that extend beyond accuracy to trust, engagement, and longitudinal outcomes. It also analyzes risks including over-reliance, hallucination, bias, privacy exposure, and unequal access, and proposes future directions spanning unified evaluation, hybrid human-AI models, memory architectures, cross-domain benchmarking, and governance. The goal is to support responsible integration of LLMs in sensitive settings where people need accompaniment and guidance, not only answers.
Authors: Joe Shymanski, Jacob Brue, Sandip Sen
Abstract: Explainable Artificial Intelligence (XAI) aims to create transparency in modern AI models by offering explanations of the models to human users. There are many ways in which researchers have attempted to evaluate the quality of these XAI models, such as user studies or proposed objective metrics like "fidelity". However, these current XAI evaluation techniques are ad hoc at best and not generalizable. Thus, most studies done within this field conduct simple user surveys to analyze the difference between no explanations and those generated by their proposed solution. We do not find this to provide adequate evidence that the explanations generated are of good quality since we believe any kind of explanation will be "better" in most metrics when compared to none at all. Thus, our study looks to highlight this pitfall: most explanations, regardless of quality or correctness, will increase user satisfaction. We also propose that emphasis should be placed on actionable explanations. We demonstrate the validity of both of our claims using an agent assistant to teach chess concepts to users. The results of this chapter will act as a call to action in the field of XAI for more comprehensive evaluation techniques for future research in order to prove explanation quality beyond user satisfaction. Additionally, we present an analysis of the scenarios in which placebic or actionable explanations would be most useful.
Authors: Fengming Liu, Shubin Yu
Abstract: We present MimiTalk, a dual-agent constitutional AI framework designed for scalable and ethical conversational data collection in social science research. The framework integrates a supervisor model for strategic oversight and a conversational model for question generation. We conducted three studies: Study 1 evaluated usability with 20 participants; Study 2 compared 121 AI interviews to 1,271 human interviews from the MediaSum dataset using NLP metrics and propensity score matching; Study 3 involved 10 interdisciplinary researchers conducting both human and AI interviews, followed by blind thematic analysis. Results across studies indicate that MimiTalk reduces interview anxiety, maintains conversational coherence, and outperforms human interviews in information richness, coherence, and stability. AI interviews elicit technical insights and candid views on sensitive topics, while human interviews better capture cultural and emotional nuances. These findings suggest that dual-agent constitutional AI supports effective human-AI collaboration, enabling replicable, scalable and quality-controlled qualitative research.
Authors: Hans Schumann, Louis Rosenberg, Ganesh Mani, Gregg Willcox
Abstract: Hyperchat AI is a novel agentic technology that enables thoughtful conversations among networked human groups of potentially unlimited size. It allows large teams to discuss complex issues, brainstorm ideas, surface risks, assess alternatives and efficiently converge on optimized solutions that amplify the group's Collective Intelligence (CI). A formal study was conducted to quantify the forecasting accuracy of human groups using Hyperchat AI to conversationally predict the outcome of Major League Baseball (MLB) games. During an 8-week period, networked groups of approximately 24 sports fans were tasked with collaboratively forecasting the winners of 59 baseball games through real-time conversation facilitated by AI agents. The results showed that when debating the games using Hyperchat AI technology, the groups converged on High Confidence predictions that significantly outperformed Vegas betting markets. Specifically, groups were 78% accurate in their High Confidence picks, a statistically strong result vs the Vegas odds of 57% (p=0.020). Had the groups bet against the spread (ATS) on these games, they would have achieved a 46% ROI against Vegas betting markets. In addition, High Confidence forecasts that were generated through above-average conversation rates were 88% accurate, suggesting that real-time interactive deliberation is central to amplified accuracy.
Authors: Marios Impraimakis
Abstract: The response-only model class selection capability of a novel deep convolutional neural network method is examined herein in a simple, yet effective, manner. Specifically, the responses from a unique degree of freedom along with their class information train and validate a one-dimensional convolutional neural network. In doing so, the network selects the model class of new and unlabeled signals without the need of the system input information, or full system identification. An optional physics-based algorithm enhancement is also examined using the Kalman filter to fuse the system response signals using the kinematics constraints of the acceleration and displacement data. Importantly, the method is shown to select the model class in slight signal variations attributed to the damping behavior or hysteresis behavior on both linear and nonlinear dynamic systems, as well as on a 3D building finite element model, providing a powerful tool for structural health monitoring applications.
Authors: Ali Safa, Farida Mohsen, Zainab Ali, Bo Wang, Amine Bermak
Abstract: Memristive crossbars enable in-memory multiply-accumulate and local plasticity learning, offering a path to energy-efficient edge AI. To this end, we present Open-MENA (Open Memristor-in-Memory Accelerator), which, to our knowledge, is the first fully open memristor interfacing system integrating (i) a reproducible hardware interface for memristor crossbars with mixed-signal read-program-verify loops; (ii) a firmware-software stack with high-level APIs for inference and on-device learning; and (iii) a Voltage-Incremental Proportional-Integral (VIPI) method to program pre-trained weights into analog conductances, followed by chip-in-the-loop fine-tuning to mitigate device non-idealities. OpenMENA is validated on digit recognition, demonstrating the flow from weight transfer to on-device adaptation, and on a real-world robot obstacle-avoidance task, where the memristor-based model learns to map localization inputs to motor commands. OpenMENA is released as open source to democratize memristor-enabled edge-AI research.
Authors: Oluwadurotimi Onibonoje, Vuong M. Ngo, Andrew McCarre, Elodie Ruelle, Bernadette O-Briend, Mark Roantree
Abstract: Grasslands, constituting the world's second-largest terrestrial carbon sink, play a crucial role in biodiversity and the regulation of the carbon cycle. Currently, the Irish dairy sector, a significant economic contributor, grapples with challenges related to profitability and sustainability. Presently, grass growth forecasting relies on impractical mechanistic models. In response, we propose deep learning models tailored for univariate datasets, presenting cost-effective alternatives. Notably, a temporal convolutional network designed for forecasting Perennial Ryegrass growth in Cork exhibits high performance, leveraging historical grass height data with RMSE of 2.74 and MAE of 3.46. Validation across a comprehensive dataset spanning 1,757 weeks over 34 years provides insights into optimal model configurations. This study enhances our understanding of model behavior, thereby improving reliability in grass growth forecasting and contributing to the advancement of sustainable dairy farming practices.
Authors: Youssef Elmir, Yassine Himeur, Abbes Amira
Abstract: This study presents a federated learning (FL) framework for privacy-preserving electrocardiogram (ECG) classification in Internet of Things (IoT) healthcare environments. By transforming 1D ECG signals into 2D Gramian Angular Field (GAF) images, the proposed approach enables efficient feature extraction through Convolutional Neural Networks (CNNs) while ensuring that sensitive medical data remain local to each device. This work is among the first to experimentally validate GAF-based federated ECG classification across heterogeneous IoT devices, quantifying both performance and communication efficiency. To evaluate feasibility in realistic IoT settings, we deployed the framework across a server, a laptop, and a resource-constrained Raspberry Pi 4, reflecting edge-cloud integration in IoT ecosystems. Experimental results demonstrate that the FL-GAF model achieves a high classification accuracy of 95.18% in a multi-client setup, significantly outperforming a single-client baseline in both accuracy and training time. Despite the added computational complexity of GAF transformations, the framework maintains efficient resource utilization and communication overhead. These findings highlight the potential of lightweight, privacy-preserving AI for IoT-based healthcare monitoring, supporting scalable and secure edge deployments in smart health systems.
Authors: Xuan Ouyang, Senan Wang, Bouzhou Wang, Siyuan Xiahou, Jinrong Zhou, Yuekang Li
Abstract: Short-video platforms have become a central medium in the modern Internet landscape, where efficient information delivery and strong interactivity are reshaping user engagement and cultural dissemination. Among the various forms of user interaction, comments play a vital role in fostering community participation and enabling content re-creation. However, generating comments that are both compliant with platform guidelines and capable of exhibiting stylistic diversity and contextual awareness remains a significant challenge. We introduce LOLGORITHM, a modular multi-agent system (MAS) designed for controllable short-video comment generation. The system integrates video segmentation, contextual and affective analysis, and style-aware prompt construction. It supports six distinct comment styles: puns (homophones), rhyming, meme application, sarcasm (irony), plain humor, and content extraction. Powered by a multimodal large language model (MLLM), LOLGORITHM directly processes video inputs and achieves fine-grained style control through explicit prompt markers and few-shot examples. To support development and evaluation, we construct a bilingual dataset using official APIs from Douyin (Chinese) and YouTube (English), covering five popular video genres: comedy skits, daily life jokes, funny animal clips, humorous commentary, and talk shows. Evaluation combines automated metrics originality, relevance, and style conformity with a large-scale human preference study involving 40 videos and 105 participants. Results show that LOLGORITHM significantly outperforms baseline models, achieving preference rates of over 90% on Douyin and 87.55% on YouTube. This work presents a scalable and culturally adaptive framework for stylized comment generation on short-video platforms, offering a promising path to enhance user engagement and creative interaction.
Authors: Jiarui Ji, Runlin Lei, Xuchen Pan, Zhewei Wei, Hao Sun, Yankai Lin, Xu Chen, Yongzheng Yang, Yaliang Li, Bolin Ding, Ji-Rong Wen
Abstract: The emergence of Large Language Models (LLMs) demonstrates their potential to encapsulate the logic and patterns inherent in human behavior simulation by leveraging extensive web data pre-training. However, the boundaries of LLM capabilities in social simulation remain unclear. To further explore the social attributes of LLMs, we introduce the CiteAgent framework, designed to generate citation networks based on human-behavior simulation with LLM-based agents. CiteAgent successfully captures predominant phenomena in real-world citation networks, including power-law distribution, citational distortion, and shrinking diameter. Building on this realistic simulation, we establish two LLM-based research paradigms in social science: LLM-SE (LLM-based Survey Experiment) and LLM-LE (LLM-based Laboratory Experiment). These paradigms facilitate rigorous analyses of citation network phenomena, allowing us to validate and challenge existing theories. Additionally, we extend the research scope of traditional science of science studies through idealized social experiments, with the simulation experiment results providing valuable insights for real-world academic environments. Our work demonstrates the potential of LLMs for advancing science of science research in social science.
Authors: Umut \c{C}al{\i}ky{\i}lmaz, Nitin Nayak, Jinghua Groppe, Sven Groppe
Abstract: In recent years, the research of multi-agent systems has taken a direction to explore larger and more complex models to fulfill sophisticated tasks. We point out two possible pitfalls that might be caused by increasing complexity; susceptibilities to faults, and performance bottlenecks. To prevent the former threat, we propose a transaction-based framework to design very complex multi-agent systems (VCMAS). To address the second threat, we offer to integrate transaction scheduling into the proposed framework. We implemented both of these ideas to develop the OptiMA framework and show that it is able to facilitate the execution of VCMAS with more than a hundred agents. We also demonstrate the effect of transaction scheduling on such a system by showing improvements up to more than 16\%. Furthermore, we also performed a theoretical analysis on the transaction scheduling problem and provided practical tools that can be used for future research on it.
Authors: Alif Elham Khan
Abstract: Medical image labels are often organized by taxonomies (e.g., organ - tissue - subtype), yet standard self-supervised learning (SSL) ignores this structure. We present a hierarchy-preserving contrastive framework that makes the label tree a first-class training signal and an evaluation target. Our approach introduces two plug-in objectives: Hierarchy-Weighted Contrastive (HWC), which scales positive/negative pair strengths by shared ancestors to promote within-parent coherence, and Level-Aware Margin (LAM), a prototype margin that separates ancestor groups across levels. The formulation is geometry-agnostic and applies to Euclidean and hyperbolic embeddings without architectural changes. Across several benchmarks, including breast histopathology, the proposed objectives consistently improve representation quality over strong SSL baselines while better respecting the taxonomy. We evaluate with metrics tailored to hierarchy faithfulness: HF1 (hierarchical F1), H-Acc (tree-distance-weighted accuracy), and parent-distance violation rate. We also report top-1 accuracy for completeness. Ablations show that HWC and LAM are effective even without curvature, and combining them yields the most taxonomy-aligned representations. Taken together, these results provide a simple, general recipe for learning medical image representations that respect the label tree and advance both performance and interpretability in hierarchy-rich domains.
Authors: Haoyu Guo, Maria Tikhanovskaya, Paul Raccuglia, Alexey Vlaskin, Chris Co, Daniel J. Liebling, Scott Ellsworth, Matthew Abraham, Elizabeth Dorfman, N. P. Armitage, Chunhan Feng, Antoine Georges, Olivier Gingras, Dominik Kiese, Steven A. Kivelson, Vadim Oganesyan, B. J. Ramshaw, Subir Sachdev, T. Senthil, J. M. Tranquada, Michael P. Brenner, Subhashini Venugopalan, Eun-Ah Kim
Abstract: Large Language Models (LLMs) show great promise as a powerful tool for scientific literature exploration. However, their effectiveness in providing scientifically accurate and comprehensive answers to complex questions within specialized domains remains an active area of research. Using the field of high-temperature cuprates as an exemplar, we evaluate the ability of LLM systems to understand the literature at the level of an expert. We construct an expert-curated database of 1,726 scientific papers that covers the history of the field, and a set of 67 expert-formulated questions that probe deep understanding of the literature. We then evaluate six different LLM-based systems for answering these questions, including both commercially available closed models and a custom retrieval-augmented generation (RAG) system capable of retrieving images alongside text. Experts then evaluate the answers of these systems against a rubric that assesses balanced perspectives, factual comprehensiveness, succinctness, and evidentiary support. Among the six systems two using RAG on curated literature outperformed existing closed models across key metrics, particularly in providing comprehensive and well-supported answers. We discuss promising aspects of LLM performances as well as critical short-comings of all the models. The set of expert-formulated questions and the rubric will be valuable for assessing expert level performance of LLM based reasoning systems.
Authors: Bo Zhao, Berkcan Kapusuzoglu, Kartik Balasubramaniam, Sambit Sahu, Supriyo Chakraborty, Genta Indra Winata
Abstract: Reasoning language models perform well on complex tasks but are costly to deploy due to their size and long reasoning traces. We propose a routing approach that assigns each problem to the smallest model likely to solve it, reducing compute without sacrificing accuracy. Using intermediate representations from s1.1-32B, we train lightweight predictors of problem difficulty or model correctness to guide routing across a pool of reasoning models. On diverse math benchmarks, routing improves efficiency over random assignment and matches s1.1-32B's performance while using significantly less compute. Our results demonstrate that difficulty-aware routing is effective for cost-efficient deployment of reasoning models.
Authors: Jan Koco\'n, Maciej Piasecki, Arkadiusz Janz, Teddy Ferdinan, {\L}ukasz Radli\'nski, Bart{\l}omiej Koptyra, Marcin Oleksy, Stanis{\l}aw Wo\'zniak, Pawe{\l} Walkowiak, Konrad Wojtasik, Julia Moska, Tomasz Naskr\k{e}t, Bartosz Walkowiak, Mateusz Gniewkowski, Kamil Szyc, Dawid Motyka, Dawid Banach, Jonatan Dalasi\'nski, Ewa Rudnicka, Bart{\l}omiej Alberski, Tomasz Walkowiak, Aleksander Szcz\k{e}sny, Maciej Markiewicz, Tomasz Berna\'s, Hubert Mazur, Kamil \.Zyta, Mateusz Tykierko, Grzegorz Chodak, Tomasz Kajdanowicz, Przemys{\l}aw Kazienko, Agnieszka Karli\'nska, Karolina Seweryn, Anna Ko{\l}os, Maciej Chrab\k{a}szcz, Katarzyna Lorenc, Aleksandra Krasnod\k{e}bska, Artur Wilczek, Katarzyna Dziewulska, Paula Betscher, Zofia Cie\'sli\'nska, Katarzyna Kowol, Daria Miko\'s, Maciej Trzci\'nski, Dawid Krutul, Marek Koz{\l}owski, S{\l}awomir Dadas, Rafa{\l} Po\'swiata, Micha{\l} Pere{\l}kiewicz, Ma{\l}gorzata Gr\k{e}bowiec, Maciej Kazu{\l}a, Marcin Bia{\l}as, Roman Roszko, Danuta Roszko, Jurgita Vai\v{c}enonien\.e, Andrius Utka, Pawe{\l} Levchuk, Pawe{\l} Kowalski, Irena Prawdzic-Jankowska, Maciej Ogrodniczuk, Monika Borys, Anna Buli\'nska, Wiktoria Gumienna, Witold Kiera\'s, Dorota Komosi\'nska, Katarzyna Krasnowska-Kiera\'s, {\L}ukasz Kobyli\'nski, Martyna Lewandowska, Marek {\L}azi\'nski, Miko{\l}aj {\L}\k{a}tkowski, Dawid Mastalerz, Beata Milewicz, Agnieszka Anna Mykowiecka, Angelika Peljak-{\L}api\'nska, Sandra Penno, Zuzanna Przybysz, Micha{\l} Rudolf, Piotr Rybak, Karolina Saputa, Aleksandra Tomaszewska, Aleksander Wawer, Marcin Woli\'nski, Joanna Wo{\l}oszyn, Alina Wr\'oblewska, Bartosz \.Zuk, Filip \.Zarnecki, Konrad Kaczy\'nski, Anna Cichosz, Zuzanna Deckert, Monika Garnys, Izabela Grabarczyk, Wojciech Janowski, Sylwia Karasi\'nska, Aleksandra Kujawiak, Piotr Misztela, Maria Szyma\'nska, Karolina Walkusz, Igor Siek, Jakub Kwiatkowski, Piotr P\k{e}zik
Abstract: Large Language Models (LLMs) play a central role in modern artificial intelligence, yet their development has been primarily focused on English, resulting in limited support for other languages. We present PLLuM (Polish Large Language Model), the largest open-source family of foundation models tailored specifically for the Polish language. Developed by a consortium of major Polish research institutions, PLLuM addresses the need for high-quality, transparent, and culturally relevant language models beyond the English-centric commercial landscape. We describe the development process, including the construction of a new 140-billion-token Polish text corpus for pre-training, a 77k custom instructions dataset, and a 100k preference optimization dataset. A key component is a Responsible AI framework that incorporates strict data governance and a hybrid module for output correction and safety filtering. We detail the models' architecture, training procedures, and alignment techniques for both base and instruction-tuned variants, and demonstrate their utility in a downstream task within public administration. By releasing these models publicly, PLLuM aims to foster open research and strengthen sovereign AI technologies in Poland.
Authors: Esha Sadia Nasir, Behnaz Elhaminia, Mark Eastwood, Catherine King, Owen Cain, Lorraine Harper, Paul Moss, Dimitrios Chanouzas, David Snead, Nasir Rajpoot, Adam Shephard, Shan E Ahmed Raza
Abstract: Accurate and efficient registration of whole slide images (WSIs) is essential for high-resolution, nuclei-level analysis in multi-stained tissue slides. We propose a novel coarse-to-fine framework CORE for accurate nuclei-level registration across diverse multimodal whole-slide image (WSI) datasets. The coarse registration stage leverages prompt-based tissue mask extraction to effectively filter out artefacts and non-tissue regions, followed by global alignment using tissue morphology and ac- celerated dense feature matching with a pre-trained feature extractor. From the coarsely aligned slides, nuclei centroids are detected and subjected to fine-grained rigid registration using a custom, shape-aware point-set registration model. Finally, non-rigid alignment at the cellular level is achieved by estimating a non-linear dis- placement field using Coherent Point Drift (CPD). Our approach benefits from automatically generated nuclei that enhance the accuracy of deformable registra- tion and ensure precise nuclei-level correspondence across modalities. The pro- posed model is evaluated on three publicly available WSI registration datasets, and two private datasets. We show that CORE outperforms current state-of-the-art methods in terms of generalisability, precision, and robustness in bright-field and immunofluorescence microscopy WSIs
Authors: Duong Mai, Lawrence Hall
Abstract: Deep learned (DL) models for image recognition have been shown to fail to generalize to data from different devices, populations, etc. COVID-19 detection from Chest X-rays (CXRs), in particular, has been shown to fail to generalize to out-of-distribution (OOD) data from new clinical sources not covered in the training set. This occurs because models learn to exploit shortcuts - source-specific artifacts that do not translate to new distributions - rather than reasonable biomarkers to maximize performance on in-distribution (ID) data. Rendering the models more robust to distribution shifts, our study investigates the use of fundamental noise injection techniques (Gaussian, Speckle, Poisson, and Salt and Pepper) during training. Our empirical results demonstrate that this technique can significantly reduce the performance gap between ID and OOD evaluation from 0.10-0.20 to 0.01-0.06, based on results averaged over ten random seeds across key metrics such as AUC, F1, accuracy, recall and specificity. Our source code is publicly available at https://github.com/Duongmai127/Noisy-ood
Authors: Tammy Mackenzie, Sukriti Punj, Natalie Perez, Sreyoshi Bhaduri, Branislav Radeljic
Abstract: This paper examines how decision makers in academia, government, business, and civil society navigate questions of power in implementations of artificial intelligence. The study explores how individuals experience and exercise levers of power, which are presented as social mechanisms that shape institutional responses to technological change. The study reports on the responses of personalized questionnaires designed to gather insight on a decision maker's institutional purview, based on an institutional governance framework developed from the work of Neo-institutionalists. Findings present the anonymized, real responses and circumstances of respondents in the form of twelve fictional personas of high-level decision makers from North America and Europe. These personas illustrate how personal agency, organizational logics, and institutional infrastructures may intersect in the governance of AI. The decision makers' responses to the questionnaires then inform a discussion of the field-level personal power of decision makers, methods of fostering institutional stability in times of change, and methods of influencing institutional change in the field of AI. The final section of the discussion presents a table of the dynamics of the levers of power in the field of AI for change makers and five testable hypotheses for institutional and social movement researchers. In summary, this study provides insight on the means for policymakers within institutions and their counterparts in civil society to personally engage with AI governance.
Authors: Arijit Bhattacharjee, Ali TehraniJamsaz, Le Chen, Niranjan Hasabnis, Mihai Capota, Nesreen Ahmed, Ali Jannesari
Abstract: Recent advances in large language models (LLMs) have significantly accelerated progress in code translation, enabling more accurate and efficient transformation across programming languages. While originally developed for natural language processing, LLMs have shown strong capabilities in modeling programming language syntax and semantics, outperforming traditional rule-based systems in both accuracy and flexibility. These models have streamlined cross-language conversion, reduced development overhead, and accelerated legacy code migration. In this paper, we introduce OMPILOT, a novel domain-specific encoder-decoder transformer tailored for translating C++ code into OpenMP, enabling effective shared-memory parallelization. OMPILOT leverages custom pre-training objectives that incorporate the semantics of parallel constructs and combines both unsupervised and supervised learning strategies to improve code translation robustness. Unlike previous work that focused primarily on loop-level transformations, OMPILOT operates at the function level to capture a wider semantic context. To evaluate our approach, we propose OMPBLEU, a novel composite metric specifically crafted to assess the correctness and quality of OpenMP parallel constructs, addressing limitations in conventional translation metrics.
Authors: Florence Klitzner, Blanca Inigo, Benjamin D. Killeen, Lalithkumar Seenivasan, Michelle Song, Axel Krieger, Mathias Unberath
Abstract: Imitation learning-based robot control policies are enjoying renewed interest in video-based robotics. However, it remains unclear whether this approach applies to X-ray-guided procedures, such as spine instrumentation. This is because interpretation of multi-view X-rays is complex. We examine opportunities and challenges for imitation policy learning in bi-plane-guided cannula insertion. We develop an in silico sandbox for scalable, automated simulation of X-ray-guided spine procedures with a high degree of realism. We curate a dataset of correct trajectories and corresponding bi-planar X-ray sequences that emulate the stepwise alignment of providers. We then train imitation learning policies for planning and open-loop control that iteratively align a cannula solely based on visual information. This precisely controlled setup offers insights into limitations and capabilities of this method. Our policy succeeded on the first attempt in 68.5% of cases, maintaining safe intra-pedicular trajectories across diverse vertebral levels. The policy generalized to complex anatomy, including fractures, and remained robust to varied initializations. Rollouts on real bi-planar X-rays further suggest that the model can produce plausible trajectories, despite training exclusively in simulation. While these preliminary results are promising, we also identify limitations, especially in entry point precision. Full closed-look control will require additional considerations around how to provide sufficiently frequent feedback. With more robust priors and domain knowledge, such models may provide a foundation for future efforts toward lightweight and CT-free robotic intra-operative spinal navigation.
Authors: Hlali Azzeddine, Majid Ben Yakhlef, Soulaiman El Hazzat
Abstract: Small, imbalanced datasets and poor input image quality can lead to high false predictions rates with deep learning models. This paper introduces Class-Based Image Composition, an approach that allows us to reformulate training inputs through a fusion of multiple images of the same class into combined visual composites, named Composite Input Images (CoImg). That enhances the intra-class variance and improves the valuable information density per training sample and increases the ability of the model to distinguish between subtle disease patterns. Our method was evaluated on the Optical Coherence Tomography Dataset for Image-Based Deep Learning Methods (OCTDL) (Kulyabin et al., 2024), which contains 2,064 high-resolution optical coherence tomography (OCT) scans of the human retina, representing seven distinct diseases with a significant class imbalance. We constructed a perfectly class-balanced version of this dataset, named Co-OCTDL, where each scan is resented as a 3x1 layout composite image. To assess the effectiveness of this new representation, we conducted a comparative analysis between the original dataset and its variant using a VGG16 model. A fair comparison was ensured by utilizing the identical model architecture and hyperparameters for all experiments. The proposed approach markedly improved diagnostic results.The enhanced Dataset achieved near-perfect accuracy (99.6%) with F1-score (0.995) and AUC (0.9996), compared to a baseline model trained on raw dataset. The false prediction rate was also significantly lower, this demonstrates that the method can producehigh-quality predictions even for weak datasets affected by class imbalance or small sample size.
Authors: Arup Datta, Ahmed Aljohani, Hyunsook Do
Abstract: Large language models (LLMs) are now widely used to draft and refactor code, but code that works is not necessarily secure. We evaluate secure code generation using the Instruct Prime, which eliminated compliance-required prompts and cue contamination, and evaluate five instruction-tuned code LLMs using a zero-shot baseline and a three-round reflexion prompting approach. Security is measured using the Insecure Code Detector (ICD), and results are reported by measuring Repair, Regression, and NetGain metrics, considering the programming language and CWE family. Our findings show that insecurity remains common at the first round: roughly 25-33% of programs are insecure at a zero-shot baseline (t0 ). Weak cryptography/config-dependent bugs are the hardest to avoid while templated ones like XSS, code injection, and hard-coded secrets are handled more reliably. Python yields the highest secure rates; C and C# are the lowest, with Java, JS, PHP, and C++ in the middle. Reflexion prompting improves security for all models, improving average accuracy from 70.74% at t0 to 79.43% at t3 , with the largest gains in the first round followed by diminishing returns. The trends with Repair, Regression, and NetGain metrics show that applying one to two rounds produces most of the benefits. A replication package is available at https://doi.org/10.5281/zenodo.17065846.
Authors: Liam Bakar, Zachary Englhardt, Vidya Srinivas, Girish Narayanswamy, Dilini Nissanka, Shwetak Patel, Vikram Iyer
Abstract: Food logging, both self-directed and prescribed, plays a critical role in uncovering correlations between diet, medical, fitness, and health outcomes. Through conversations with nutritional experts and individuals who practice dietary tracking, we find current logging methods, such as handwritten and app-based journaling, are inflexible and result in low adherence and potentially inaccurate nutritional summaries. These findings, corroborated by prior literature, emphasize the urgent need for improved food logging methods. In response, we propose SnappyMeal, an AI-powered dietary tracking system that leverages multimodal inputs to enable users to more flexibly log their food intake. SnappyMeal introduces goal-dependent follow-up questions to intelligently seek missing context from the user and information retrieval from user grocery receipts and nutritional databases to improve accuracy. We evaluate SnappyMeal through publicly available nutrition benchmarks and a multi-user, 3-week, in-the-wild deployment capturing over 500 logged food instances. Users strongly praised the multiple available input methods and reported a strong perceived accuracy. These insights suggest that multimodal AI systems can be leveraged to significantly improve dietary tracking flexibility and context-awareness, laying the groundwork for a new class of intelligent self-tracking applications.
Authors: Nand Kumar Yadav (AI Research Lab, Department of Computer Science,Biomedical,Translational Sciences, Sanford School of Medicine, University Of South Dakota, Vermillion, SD, USA), Rodrigue Rizk (AI Research Lab, Department of Computer Science,Biomedical,Translational Sciences, Sanford School of Medicine, University Of South Dakota, Vermillion, SD, USA), William CW Chen (AI Research Lab, Department of Computer Science,Biomedical,Translational Sciences, Sanford School of Medicine, University Of South Dakota, Vermillion, SD, USA), KC Santosh (AI Research Lab, Department of Computer Science,Biomedical,Translational Sciences, Sanford School of Medicine, University Of South Dakota, Vermillion, SD, USA)
Abstract: Unknown anomaly detection in medical imaging remains a fundamental challenge due to the scarcity of labeled anomalies and the high cost of expert supervision. We introduce an unsupervised, oracle-free framework that incrementally expands a trusted set of normal samples without any anomaly labels. Starting from a small, verified seed of normal images, our method alternates between lightweight adapter updates and uncertainty-gated sample admission. A frozen pretrained vision backbone is augmented with tiny convolutional adapters, ensuring rapid domain adaptation with negligible computational overhead. Extracted embeddings are stored in a compact coreset enabling efficient k-nearest neighbor anomaly (k-NN) scoring. Safety during incremental expansion is enforced by dual probabilistic gates, a sample is admitted into the normal memory only if its distance to the existing coreset lies within a calibrated z-score threshold, and its SWAG-based epistemic uncertainty remains below a seed-calibrated bound. This mechanism prevents drift and false inclusions without relying on generative reconstruction or replay buffers. Empirically, our system steadily refines the notion of normality as unlabeled data arrive, producing substantial gains over baselines. On COVID-CXR, ROC-AUC improves from 0.9489 to 0.9982 (F1: 0.8048 to 0.9746); on Pneumonia CXR, ROC-AUC rises from 0.6834 to 0.8968; and on Brain MRI ND-5, ROC-AUC increases from 0.6041 to 0.7269 and PR-AUC from 0.7539 to 0.8211. These results highlight the effectiveness and efficiency of the proposed framework for real-world, label-scarce medical imaging applications.
Authors: Dom\'icio Pereira Neto, Jo\~ao Correia, Penousal Machado
Abstract: Deep generative models, especially diffusion architectures, have transformed image generation; however, they are challenging to control and optimize for specific goals without expensive retraining. Embedding Space Exploration, especially with Evolutionary Algorithms (EAs), has been shown to be a promising method for optimizing image generation, particularly within Diffusion Models. Therefore, in this work, we study the performance of an evolutionary optimization method, namely Separable Covariance Matrix Adaptation Evolution Strategy (sep-CMA-ES), against the widely adopted Adaptive Moment Estimation (Adam), applied to Stable Diffusion XL Turbo's prompt embedding vector. The evaluation of images combines the LAION Aesthetic Predictor V2 with CLIPScore into a weighted fitness function, allowing flexible trade-offs between visual appeal and adherence to prompts. Experiments on a subset of the Parti Prompts (P2) dataset showcase that sep-CMA-ES consistently yields superior improvements in aesthetic and alignment metrics in comparison to Adam. Results indicate that the evolutionary method provides efficient, gradient-free optimization for diffusion models, enhancing controllability without the need for fine-tuning. This study emphasizes the potential of evolutionary methods for embedding space exploration of deep generative models and outlines future research directions.
Authors: Nikta Akbarpour, Mahdieh Sadat Benis, Fatemeh Hendijani Fard, Ali Ouni, Mohamed Aymen Saied
Abstract: Automated Program Repair (APR) has advanced rapidly with Large Language Models (LLMs), but most existing methods remain computationally expensive, and focused on a small set of languages. Ruby, despite its widespread use in web development and the persistent challenges faced by its developers, has received little attention in APR research. In this paper, we introduce RAMP, a novel lightweight framework that formulates program repair as a feedback-driven, iterative process for Ruby. RAMP employs a team of collaborative agents that generate targeted tests, reflect on errors, and refine candidate fixes until a correct solution is found. Unlike prior approaches, RAMP is designed to avoid reliance on large multilingual repair databases or costly fine-tuning, instead operating directly on Ruby through lightweight prompting and test-driven feedback. Evaluation on the XCodeEval benchmark shows that RAMP achieves a pass@1 of 67% on Ruby, outper-forming prior approaches. RAMP converges quickly within five iterations, and ablation studies confirm that test generation and self-reflection are key drivers of its performance. Further analysis shows that RAMP is particularly effective at repairing wrong answers, compilation errors, and runtime errors. Our approach provides new insights into multi-agent repair strategies, and establishes a foundation for extending LLM-based debugging tools to under-studied languages.
Authors: NVIDIA (Danny), : (Danny), Amala Sanjay Deshmukh (Danny), Kateryna Chumachenko (Danny), Tuomas Rintamaki (Danny), Matthieu Le (Danny), Tyler Poon (Danny), Danial Mohseni Taheri (Danny), Ilia Karmanov (Danny), Guilin Liu (Danny), Jarno Seppanen (Danny), Guo Chen (Danny), Karan Sapra (Danny), Zhiding Yu (Danny), Adi Renduchintala (Danny), Charles Wang (Danny), Peter Jin (Danny), Arushi Goel (Danny), Mike Ranzinger (Danny), Lukas Voegtle (Danny), Philipp Fischer (Danny), Timo Roman (Danny), Wei Ping (Danny), Boxin Wang (Danny), Zhuolin Yang (Danny), Nayeon Lee (Danny), Shaokun Zhang (Danny), Fuxiao Liu (Danny), Zhiqi Li (Danny), Di Zhang (Danny), Greg Heinrich (Danny), Hongxu (Danny), Yin (Justin), Song Han (Justin), Pavlo Molchanov (Justin), Parth Mannan (Justin), Yao Xu (Justin), Jane Polak Scowcroft (Justin), Tom Balough (Justin), Subhashree Radhakrishnan (Justin), Paris Zhang (Justin), Sean Cha (Justin), Ratnesh Kumar (Justin), Zaid Pervaiz Bhat (Justin), Jian Zhang (Justin), Darragh Hanley (Justin), Pritam Biswas (Justin), Jesse Oliver (Justin), Kevin Vasques (Justin), Roger Waleffe (Justin), Duncan Riach (Justin), Oluwatobi Olabiyi (Justin), Ameya Sunil Mahabaleshwarkar (Justin), Bilal Kartal (Justin), Pritam Gundecha (Justin), Khanh Nguyen (Justin), Alexandre Milesi (Justin), Eugene Khvedchenia (Justin), Ran Zilberstein (Justin), Ofri Masad (Justin), Natan Bagrov (Justin), Nave Assaf (Justin), Tomer Asida (Justin), Daniel Afrimi (Justin), Amit Zuker (Justin), Netanel Haber (Justin), Zhiyu Cheng (Justin), Jingyu (Justin), Xin (Allan), Di (Allan), Wu, Nik Spirin, Maryam Moosaei, Roman Ageev, Vanshil Atul Shah, Yuting Wu, Daniel Korzekwa, Unnikrishnan Kizhakkemadam Sreekumar, Wanli Jiang, Padmavathy Subramanian, Alejandra Rico, Sandip Bhaskar, Saeid Motiian, Kedi Wu, Annie Surla, Chia-Chih Chen, Hayden Wolff, Matthew Feinberg, Melissa Corpuz, Marek Wawrzos, Eileen Long, Aastha Jhunjhunwala, Paul Hendricks, Farzan Memarian, Benika Hall, Xin-Yu Wang, David Mosallanezhad, Soumye Singhal, Luis Vega, Katherine Cheung, Krzysztof Pawelec, Michael Evans, Katherine Luna, Jie Lou, Erick Galinkin, Akshay Hazare, Kaustubh Purandare, Ann Guan, Anna Warno, Chen Cui, Yoshi Suhara, Shibani Likhite, Seph Mard, Meredith Price, Laya Sleiman, Saori Kaji, Udi Karpas, Kari Briski, Joey Conway, Michael Lightstone, Jan Kautz, Mohammad Shoeybi, Mostofa Patwary, Jonathen Cohen, Oleksii Kuchaiev, Andrew Tao, Bryan Catanzaro
Abstract: We introduce Nemotron Nano V2 VL, the latest model of the Nemotron vision-language series designed for strong real-world document understanding, long video comprehension, and reasoning tasks. Nemotron Nano V2 VL delivers significant improvements over our previous model, Llama-3.1-Nemotron-Nano-VL-8B, across all vision and text domains through major enhancements in model architecture, datasets, and training recipes. Nemotron Nano V2 VL builds on Nemotron Nano V2, a hybrid Mamba-Transformer LLM, and innovative token reduction techniques to achieve higher inference throughput in long document and video scenarios. We are releasing model checkpoints in BF16, FP8, and FP4 formats and sharing large parts of our datasets, recipes and training code.
Authors: Athma Narayanan, Mahesh Subedar, Omesh Tickoo
Abstract: We present an agentic flow consisting of multiple agents that combine specialized LLMs and hardware simulation tools to collaboratively complete the complex task of Register Transfer Level (RTL) generation without human intervention. A key feature of the proposed flow is the progressive error feedback system of agents (PEFA), a self-correcting mechanism that leverages iterative error feedback to progressively increase the complexity of the approach. The generated RTL includes checks for compilation, functional correctness, and synthesizable constructs. To validate this adaptive approach to code generation, benchmarking is performed using two opensource natural language-to-RTL datasets. We demonstrate the benefits of the proposed approach implemented on an open source agentic framework, using both open- and closed-source LLMs, effectively bridging the performance gap between them. Compared to previously published methods, our approach sets a new benchmark, providing state-of-the-art pass rates while being efficient in token counts.
Authors: Raghav Sharma, Manan Mehta, Sai Tiger Raina
Abstract: Reinforcement Learning from Human Feedback (RLHF) is the standard for aligning Large Language Models (LLMs), yet recent progress has moved beyond canonical text-based methods. This survey synthesizes the new frontier of alignment research by addressing critical gaps in multi-modal alignment, cultural fairness, and low-latency optimization. To systematically explore these domains, we first review foundational algo- rithms, including PPO, DPO, and GRPO, before presenting a detailed analysis of the latest innovations. By providing a comparative synthesis of these techniques and outlining open challenges, this work serves as an essential roadmap for researchers building more robust, efficient, and equitable AI systems.
Authors: Fu-Chun Yang, Jason Eshraghian
Abstract: In multi-agent settings, such as debate, reflection, or tool-calling, large language models (LLMs) pass messages as plain tokens, discarding most latent semantics. This constrains information transfer and adds unnecessary computational overhead. We form a latent bridge via vector translations, which use learned mappings that enable direct semantic exchange between representation spaces. A dual-encoder translator trained between Llama-2-7B and Mistral-7B-Instruct attains an average cosine alignment of 0.538. Injecting the translated vectors at 30 percent blending strength steers the target model's generation without destabilizing logits. Bidirectional evaluation shows a 2.01:1 transfer asymmetry, indicating that general-purpose models yield more transferable representations than instruction-tuned variants. This conservative injection preserves computational stability while demonstrating that cross-model latent communication is feasible, enabling collaborative AI systems that share meaning rather than tokens.
Authors: Zhejia Cai, Puhua Jiang, Shiwei Mao, Hongkun Cao, Ruqi Huang
Abstract: Reconstructing real-world objects from multi-view images is essential for applications in 3D editing, AR/VR, and digital content creation. Existing methods typically prioritize either geometric accuracy (Multi-View Stereo) or photorealistic rendering (Novel View Synthesis), often decoupling geometry and appearance optimization, which hinders downstream editing tasks. This paper advocates an unified treatment on geometry and appearance optimization for seamless Gaussian-mesh joint optimization. More specifically, we propose a novel framework that simultaneously optimizes mesh geometry (vertex positions and faces) and vertex colors via Gaussian-guided mesh differentiable rendering, leveraging photometric consistency from input images and geometric regularization from normal and depth maps. The obtained high-quality 3D reconstruction can be further exploit in down-stream editing tasks, such as relighting and shape deformation. The code will be publicly available upon acceptance.
Authors: Xu Zou
Abstract: Since its emergence, SARS-CoV-2 has demonstrated a rapid and unpredictable evolutionary trajectory, characterized by the continual emergence of immune-evasive variants. This poses persistent challenges to public health and vaccine development. While large-scale generative pre-trained transformers (GPTs) have revolutionized the modeling of sequential data, their direct applications to noisy viral genomic sequences are limited. In this paper, we introduce PETRA(Pretrained Evolutionary TRAnsformer), a novel transformer approach based on evolutionary trajectories derived from phylogenetic trees rather than raw RNA sequences. This method effectively mitigates sequencing noise and captures the hierarchical structure of viral evolution. With a weighted training framework to address substantial geographical and temporal imbalances in global sequence data, PETRA excels in predicting future SARS-CoV-2 mutations, achieving a weighted recall@1 of 9.45% for nucleotide mutations and 17.10\% for spike amino-acid mutations, compared to 0.49% and 6.64% respectively for the best baseline. PETRA also demonstrates its ability to aid in the real-time mutation prediction of major clades like 24F(XEC) and 25A(LP.8.1). The code is open sourced on https://github.com/xz-keg/PETra
Authors: Berk Iskar, Michael Taynnan Barros
Abstract: Network anomaly detection systems encounter several challenges with traditional detectors trained offline. They become susceptible to concept drift and new threats such as zero-day or polymorphic attacks. To address this limitation, we propose a Ca$^{2+}$-modulated learning framework that draws inspiration from astrocytic Ca$^{2+}$ signaling in the brain, where rapid, context-sensitive adaptation enables robust information processing. Our approach couples a multicellular astrocyte dynamics simulator with a deep neural network (DNN). The simulator models astrocytic Ca$^{2+}$ dynamics through three key mechanisms: IP$_3$-mediated Ca$^{2+}$ release, SERCA pump uptake, and conductance-aware diffusion through gap junctions between cells. Evaluation of our proposed network on CTU-13 (Neris) network traffic data demonstrates the effectiveness of our biologically plausible approach. The Ca$^{2+}$-gated model outperforms a matched baseline DNN, achieving up to $\sim$98\% accuracy with reduced false positives and negatives across multiple train/test splits. Importantly, this improved performance comes with negligible runtime overhead once Ca$^{2+}$ trajectories are precomputed. While demonstrated here for cybersecurity applications, this Ca$^{2+}$-modulated learning framework offers a generic solution for streaming detection tasks that require rapid, biologically grounded adaptation to evolving data patterns.
Authors: Shiyin Lin
Abstract: Software fuzzing has become a cornerstone in automated vulnerability discovery, yet existing mutation strategies often lack semantic awareness, leading to redundant test cases and slow exploration of deep program states. In this work, I present a hybrid fuzzing framework that integrates static and dynamic analysis with Large Language Model (LLM)-guided input mutation and semantic feedback. Static analysis extracts control-flow and data-flow information, which is transformed into structured prompts for the LLM to generate syntactically valid and semantically diverse inputs. During execution, I augment traditional coverage-based feedback with semantic feedback signals-derived from program state changes, exception types, and output semantics-allowing the fuzzer to prioritize inputs that trigger novel program behaviors beyond mere code coverage. I implement our approach atop AFL++, combining program instrumentation with embedding-based semantic similarity metrics to guide seed selection. Evaluation on real-world open-source targets, including libpng, tcpdump, and sqlite, demonstrates that our method achieves faster time-to-first-bug, higher semantic diversity, and a competitive number of unique bugs compared to state-of-the-art fuzzers. This work highlights the potential of combining LLM reasoning with semantic-aware feedback to accelerate and deepen vulnerability discovery.
Authors: Kyaw Hpone Myint, Zhe Wu, Alexandre G. R. Day, Giri Iyengar
Abstract: Decision trees are widely used in high-stakes fields like finance and healthcare due to their interpretability. This work introduces an efficient, scalable method for generating synthetic pre-training data to enable meta-learning of decision trees. Our approach samples near-optimal decision trees synthetically, creating large-scale, realistic datasets. Using the MetaTree transformer architecture, we demonstrate that this method achieves performance comparable to pre-training on real-world data or with computationally expensive optimal decision trees. This strategy significantly reduces computational costs, enhances data generation flexibility, and paves the way for scalable and efficient meta-learning of interpretable decision tree models.
Authors: J. Nathan Kutz, Peter Battaglia, Michael Brenner, Kevin Carlberg, Aric Hagberg, Shirley Ho, Stephan Hoyer, Henning Lange, Hod Lipson, Michael W. Mahoney, Frank Noe, Max Welling, Laure Zanna, Francis Zhu, Steven L. Brunton
Abstract: Machine learning (ML) and artificial intelligence (AI) algorithms are transforming and empowering the characterization and control of dynamic systems in the engineering, physical, and biological sciences. These emerging modeling paradigms require comparative metrics to evaluate a diverse set of scientific objectives, including forecasting, state reconstruction, generalization, and control, while also considering limited data scenarios and noisy measurements. We introduce a common task framework (CTF) for science and engineering, which features a growing collection of challenge data sets with a diverse set of practical and common objectives. The CTF is a critically enabling technology that has contributed to the rapid advance of ML/AI algorithms in traditional applications such as speech recognition, language processing, and computer vision. There is a critical need for the objective metrics of a CTF to compare the diverse algorithms being rapidly developed and deployed in practice today across science and engineering.
Authors: Mingyu Sung, Vikas Palakonda, Suhwan Im, Sunghwan Moon, Il-Min Kim, Sangseok Yun, Jae-Mo Kang
Abstract: Large language models (LLMs) have achieved near-human performance across diverse reasoning tasks, yet their deployment on resource-constrained Internet-of-Things (IoT) devices remains impractical due to massive parameter footprints and memory-intensive autoregressive decoding. While split computing offers a promising solution by partitioning model execution between edge devices and cloud servers, existing approaches fail to address the unique challenges of autoregressive inference, particularly the iterative token generation process and expanding key-value (KV) cache requirements. This work introduces the first autoregressive-aware split computing framework designed explicitly for LLM deployment on edge devices. Our approach makes three key contributions. First, we develop one-point split compression (OPSC), a mixed-precision quantization scheme that prevents out-of-memory failures by strategically partitioning models into front-end and back-end segments with different precision levels. Second, we propose a two-stage intermediate compression pipeline that combines threshold splitting (TS) and token-wise adaptive bit quantization (TAB-Q) to preserve accuracy-critical activations while dramatically reducing communication overhead. Third, we formulate a unified optimization framework that jointly selects optimal split points, quantization settings, and sequence lengths to satisfy strict memory and latency constraints. Extensive evaluations across diverse LLMs and hardware platforms demonstrate superior performance compared to state-of-the-art quantization methods, including SmoothQuant, OmniQuant, and Atom. The framework achieves a 1.49 inference speedup and significant communication overhead reduction while maintaining or improving model accuracy.
Authors: Shiyin Lin
Abstract: Large Language Models (LLMs) enhanced with retrieval -- commonly referred to as Retrieval-Augmented Generation (RAG) -- have demonstrated strong performance in knowledge-intensive tasks. However, RAG pipelines often fail when retrieved evidence is incomplete, leaving gaps in the reasoning process. In such cases, \emph{abductive inference} -- the process of generating plausible missing premises to explain observations -- offers a principled approach to bridge these gaps. In this paper, we propose a framework that integrates abductive inference into retrieval-augmented LLMs. Our method detects insufficient evidence, generates candidate missing premises, and validates them through consistency and plausibility checks. Experimental results on abductive reasoning and multi-hop QA benchmarks show that our approach improves both answer accuracy and reasoning faithfulness. This work highlights abductive inference as a promising direction for enhancing the robustness and explainability of RAG systems.
Authors: Kailun Ji (School of Electronics and Information, Northwestern Polytechnical University, Xi'an, China), Xiaoyu Hu (School of Electronics and Information, Northwestern Polytechnical University, Xi'an, China), Xinyu Zhang (School of Electronics and Information, Northwestern Polytechnical University, Xi'an, China, Chongqing Institute for Brain and Intelligence, Guangyang Bay Laboratory, Chongqing, China), Jun Chen (School of Electronics and Information, Northwestern Polytechnical University, Xi'an, China, Chongqing Institute for Brain and Intelligence, Guangyang Bay Laboratory, Chongqing, China)
Abstract: Large-scale disaster Search And Rescue (SAR) operations are persistently challenged by complex terrain and disrupted communications. While Unmanned Aerial Vehicle (UAV) swarms offer a promising solution for tasks like wide-area search and supply delivery, yet their effective coordination places a significant cognitive burden on human operators. The core human-machine collaboration bottleneck lies in the ``intention-to-action gap'', which is an error-prone process of translating a high-level rescue objective into a low-level swarm command under high intensity and pressure. To bridge this gap, this study proposes a novel LLM-CRF system that leverages Large Language Models (LLMs) to model and augment human-swarm teaming cognition. The proposed framework initially captures the operator's intention through natural and multi-modal interactions with the device via voice or graphical annotations. It then employs the LLM as a cognitive engine to perform intention comprehension, hierarchical task decomposition, and mission planning for the UAV swarm. This closed-loop framework enables the swarm to act as a proactive partner, providing active feedback in real-time while reducing the need for manual monitoring and control, which considerably advances the efficacy of the SAR task. We evaluate the proposed framework in a simulated SAR scenario. Experimental results demonstrate that, compared to traditional order and command-based interfaces, the proposed LLM-driven approach reduced task completion time by approximately $64.2\%$ and improved task success rate by $7\%$. It also leads to a considerable reduction in subjective cognitive workload, with NASA-TLX scores dropping by $42.9\%$. This work establishes the potential of LLMs to create more intuitive and effective human-swarm collaborations in high-stakes scenarios.
Authors: Fatemeh Hosseinabadi, Seyedhassan Sharifi
Abstract: Pediatric appendicitis remains one of the most common causes of acute abdominal pain in children, and its diagnosis continues to challenge clinicians due to overlapping symptoms and variable imaging quality. This study aims to develop and evaluate a deep learning model based on a pretrained ResNet architecture for automated detection of appendicitis from ultrasound images. We used the Regensburg Pediatric Appendicitis Dataset, which includes ultrasound scans, laboratory data, and clinical scores from pediatric patients admitted with abdominal pain to Children Hospital. Hedwig in Regensburg, Germany. Each subject had 1 to 15 ultrasound views covering the right lower quadrant, appendix, lymph nodes, and related structures. For the image based classification task, ResNet was fine tuned to distinguish appendicitis from non-appendicitis cases. Images were preprocessed by normalization, resizing, and augmentation to enhance generalization. The proposed ResNet model achieved an overall accuracy of 93.44, precision of 91.53, and recall of 89.8, demonstrating strong performance in identifying appendicitis across heterogeneous ultrasound views. The model effectively learned discriminative spatial features, overcoming challenges posed by low contrast, speckle noise, and anatomical variability in pediatric imaging.
Authors: Fatemeh Hosseinabadi, Seyedhassan Sharifi
Abstract: Accurate segmentation of the left atrium (LA) from cardiac MRI is critical for guiding atrial fibrillation (AF) ablation and constructing biophysical cardiac models. Manual delineation is time-consuming, observer-dependent, and impractical for large-scale or time-sensitive clinical workflows. Deep learning methods, particularly convolutional architectures, have recently demonstrated superior performance in medical image segmentation tasks. In this study, we applied the nnU-Net framework, an automated, self-configuring deep learning segmentation architecture, to the Left Atrial Segmentation Challenge 2013 dataset. The dataset consists of thirty MRI scans with corresponding expert-annotated masks. The nnU-Net model automatically adapted its preprocessing, network configuration, and training pipeline to the characteristics of the MRI data. Model performance was quantitatively evaluated using the Dice similarity coefficient (DSC), and qualitative results were compared against expert segmentations. The proposed nnUNet model achieved a mean Dice score of 93.5, demonstrating high overlap with expert annotations and outperforming several traditional segmentation approaches reported in previous studies. The network exhibited robust generalization across variations in left atrial shape, contrast, and image quality, accurately delineating both the atrial body and proximal pulmonary veins.
Authors: Qingfeng Chen, Haojin Zeng, Jingyi Jie, Shichao Zhang, Debo Cheng
Abstract: With the rapid growth of graph-structured data in critical domains, unsupervised graph-level anomaly detection (UGAD) has become a pivotal task. UGAD seeks to identify entire graphs that deviate from normal behavioral patterns. However, most Graph Neural Network (GNN) approaches implicitly assume that the training set is clean, containing only normal graphs, which is rarely true in practice. Even modest contamination by anomalous graphs can distort learned representations and sharply degrade performance. To address this challenge, we propose DeNoise, a robust UGAD framework explicitly designed for contaminated training data. It jointly optimizes a graph-level encoder, an attribute decoder, and a structure decoder via an adversarial objective to learn noise-resistant embeddings. Further, DeNoise introduces an encoder anchor-alignment denoising mechanism that fuses high-information node embeddings from normal graphs into all graph embeddings, improving representation quality while suppressing anomaly interference. A contrastive learning component then compacts normal graph embeddings and repels anomalous ones in the latent space. Extensive experiments on eight real-world datasets demonstrate that DeNoise consistently learns reliable graph-level representations under varying noise intensities and significantly outperforms state-of-the-art UGAD baselines.
Authors: Brigitte A. Mora-Reyes, Jennifer A. Drewyor, Abel A. Reyes-Angulo
Abstract: Artificial intelligence (AI) systems often reflect biases from economically advanced regions, marginalizing contexts in economically developing regions like Latin America due to imbalanced datasets. This paper examines AI representations of diverse Latin American contexts, revealing disparities between data from economically advanced and developing regions. We highlight how the dominance of English over Spanish, Portuguese, and indigenous languages such as Quechua and Nahuatl perpetuates biases, framing Latin American perspectives through a Western lens. To address this, we introduce a culturally aware dataset rooted in Latin American history and socio-political contexts, challenging Eurocentric models. We evaluate six language models on questions testing cultural context awareness, using a novel Cultural Expressiveness metric, statistical tests, and linguistic analyses. Our findings show that some models better capture Latin American perspectives, while others exhibit significant sentiment misalignment (p < 0.001). Fine-tuning Mistral-7B with our dataset improves its cultural expressiveness by 42.9%, advancing equitable AI development. We advocate for equitable AI by prioritizing datasets that reflect Latin American history, indigenous knowledge, and diverse languages, while emphasizing community-centered approaches to amplify marginalized voices.
Authors: Yang Xu, Peiyao Liu, Shuwei Chen, Jun Liu
Abstract: Currently, there is a lack of rigorous theoretical system for systematically generating non-trivial and logically valid theorems. Addressing this critical gap, this paper conducts research to propose a novel automated theorem generation theory and tool. Based on the concept of standard contradiction which possesses unique deductive advantages, this paper defines and proves, for the first time, a new logical structure known as rectangular standard contradiction. Centered on this structure, a complete Automated Theorem Generation (ATG) theory is put forward. Theoretical proofs clarify two core properties of rectangular standard contradiction: first, it is a standard contradiction (necessarily unsatisfiable); second, it exhibits non-redundancy (the remaining clause set becomes satisfiable after removing any clause). Leveraging these properties, this paper proves that partitioning a rectangular standard contradiction into a premise subset $A$ and negation of its complement $H$, a valid theorem $A \vdash \neg H$ can be formed, and all such theorems are logically equivalent. To implement this theory, an efficient template-based ATG algorithm is designed, and a Rectangular Automated Theorem Generator is developed. This research enables machines to transition from "verifiers" to "discoverers", opening up new avenues for fundamental research in the fields of logic and artificial intelligence.
Authors: Moses Charikar, Chirag Pabbaraju, Ambuj Tewari
Abstract: We study the problem of language identification in the limit, where given a sequence of examples from a target language, the goal of the learner is to output a sequence of guesses for the target language such that all the guesses beyond some finite time are correct. Classical results of Gold showed that language identification in the limit is impossible for essentially any interesting collection of languages. Later, Angluin gave a precise characterization of language collections for which this task is possible. Motivated by recent positive results for the related problem of language generation, we revisit the classic language identification problem in the setting where the learner is given the additional power of producing a list of $k$ guesses at each time step. The goal is to ensure that beyond some finite time, one of the guesses is correct at each time step. We give an exact characterization of collections of languages that can be $k$-list identified in the limit, based on a recursive version of Angluin's characterization (for language identification with a list of size $1$). This further leads to a conceptually appealing characterization: A language collection can be $k$-list identified in the limit if and only if the collection can be decomposed into $k$ collections of languages, each of which can be identified in the limit (with a list of size $1$). We also use our characterization to establish rates for list identification in the statistical setting where the input is drawn as an i.i.d. stream from a distribution supported on some language in the collection. Our results show that if a collection is $k$-list identifiable in the limit, then the collection can be $k$-list identified at an exponential rate, and this is best possible. On the other hand, if a collection is not $k$-list identifiable in the limit, then it cannot be $k$-list identified at any rate that goes to zero.
Authors: Paul Badu Yakubu, Lesther Santana, Mohamed Rahouti, Yufeng Xin, Abdellah Chehri, Mohammed Aledhari
Abstract: With the increasing frequency and sophistication of Distributed Denial of Service (DDoS) attacks, it has become critical to develop more efficient and interpretable detection methods. Traditional detection systems often struggle with scalability and transparency, hindering real-time response and understanding of attack vectors. This paper presents an automated framework for detecting and interpreting DDoS attacks using machine learning (ML). The proposed method leverages the Tree-based Pipeline Optimization Tool (TPOT) to automate the selection and optimization of ML models and features, reducing the need for manual experimentation. SHapley Additive exPlanations (SHAP) is incorporated to enhance model interpretability, providing detailed insights into the contribution of individual features to the detection process. By combining TPOT's automated pipeline selection with SHAP interpretability, this approach improves the accuracy and transparency of DDoS detection. Experimental results demonstrate that key features such as mean backward packet length and minimum forward packet header length are critical in detecting DDoS attacks, offering a scalable and explainable cybersecurity solution.
Authors: Venkata Manikanta Desu, Syed Fawaz Ali
Abstract: This study presents a complete pipeline for automated tennis match analysis. Our framework integrates multiple deep learning models to detect and track players and the tennis ball in real time, while also identifying court keypoints for spatial reference. Using YOLOv8 for player detection, a custom-trained YOLOv5 model for ball tracking, and a ResNet50-based architecture for court keypoint detection, our system provides detailed analytics including player movement patterns, ball speed, shot accuracy, and player reaction times. The experimental results demonstrate robust performance in varying court conditions and match scenarios. The model outputs an annotated video along with detailed performance metrics, enabling coaches, broadcasters, and players to gain actionable insights into the dynamics of the game.
Authors: Shengyu Tang, Zeyuan Lu, Jiazhi Dong, Changdong Yu, Xiaoyu Wang, Yaohui Lyu, Weihao Xia
Abstract: Accurate perception of the marine environment through robust multi-object tracking (MOT) is essential for ensuring safe vessel navigation and effective maritime surveillance. However, the complicated maritime environment often causes camera motion and subsequent visual degradation, posing significant challenges to MOT. To address this challenge, we propose an efficient Dual-branch Maritime SORT (DMSORT) method for maritime MOT. The core of the framework is a parallel tracker with affine compensation, which incorporates an object detection and re-identification (ReID) branch, along with a dedicated branch for dynamic camera motion estimation. Specifically, a Reversible Columnar Detection Network (RCDN) is integrated into the detection module to leverage multi-level visual features for robust object detection. Furthermore, a lightweight Transformer-based appearance extractor (Li-TAE) is designed to capture global contextual information and generate robust appearance features. Another branch decouples platform-induced and target-intrinsic motion by constructing a projective transformation, applying platform-motion compensation within the Kalman filter, and thereby stabilizing true object trajectories. Finally, a clustering-optimized feature fusion module effectively combines motion and appearance cues to ensure identity consistency under noise, occlusion, and drift. Extensive evaluations on the Singapore Maritime Dataset demonstrate that DMSORT achieves state-of-the-art performance. Notably, DMSORT attains the fastest runtime among existing ReID-based MOT frameworks while maintaining high identity consistency and robustness to jitter and occlusion. Code is available at: https://github.com/BiscuitsLzy/DMSORT-An-efficient-parallel-maritime-multi-object-tracking-architecture-.
Authors: Yujian Liu, Ze Wang, Hao Chen, Ximeng Sun, Xiaodong Yu, Jialian Wu, Jiang Liu, Emad Barsoum, Zicheng Liu, Shiyu Chang
Abstract: Computer-use agents can operate computers and automate laborious tasks, but despite recent rapid progress, they still lag behind human users, especially when tasks require domain-specific procedural knowledge about particular applications, platforms, and multi-step workflows. Humans can bridge this gap by watching video tutorials: we search, skim, and selectively imitate short segments that match our current subgoal. In this paper, we study how to enable computer-use agents to learn from online videos at inference time effectively. We propose a framework that retrieves and filters tutorial videos, converts them into structured demonstration trajectories, and dynamically selects trajectories as in-context guidance during execution. Particularly, using a VLM, we infer UI actions, segment videos into short subsequences of actions, and assign each subsequence a textual objective. At inference time, a two-stage selection mechanism dynamically chooses a single trajectory to add in context at each step, focusing the agent on the most helpful local guidance for its next decision. Experiments on two widely used benchmarks show that our framework consistently outperforms strong base agents and variants that use only textual tutorials or transcripts. Analyses highlight the importance of trajectory segmentation and selection, action filtering, and visual information, suggesting that abundant online videos can be systematically distilled into actionable guidance that improves computer-use agents at inference time. Our code is available at https://github.com/UCSB-NLP-Chang/video_demo.
Authors: Boxuan Ma, Huiyong Li, Gen Li, Li Chen, Cheng Tang, Yinjie Xie, Chenghao Gu, Atsushi Shimada, Shin'ichi Konomi
Abstract: Generative AI tools such as ChatGPT now provide novice programmers with unprecedented access to instant, personalized support. While this holds clear promise, their influence on students' metacognitive processes remains underexplored. Existing work has largely focused on correctness and usability, with limited attention to whether and how students' use of AI assistants supports or bypasses key metacognitive processes. This study addresses that gap by analyzing student-AI interactions through a metacognitive lens in university-level programming courses. We examined more than 10,000 dialogue logs collected over three years, complemented by surveys of students and educators. Our analysis focused on how prompts and responses aligned with metacognitive phases and strategies. Synthesizing these findings across data sources, we distill design considerations for AI-powered coding assistants that aim to support rather than supplant metacognitive engagement. Our findings provide guidance for developing educational AI tools that strengthen students' learning processes in programming education.
Authors: Fahim Ahmed, Md Mubtasim Ahasan, Jahir Sadik Monon, Muntasir Wahed, M Ashraful Amin, A K M Mahbubur Rahman, Amin Ahsan Ali
Abstract: Text-to-SQL systems provide a natural language interface that can enable even laymen to access information stored in databases. However, existing Large Language Models (LLM) struggle with SQL generation from natural instructions due to large schema sizes and complex reasoning. Prior work often focuses on complex, somewhat impractical pipelines using flagship models, while smaller, efficient models remain overlooked. In this work, we explore three multi-agent LLM pipelines, with systematic performance benchmarking across a range of small to large open-source models: (1) Multi-agent discussion pipeline, where agents iteratively critique and refine SQL queries, and a judge synthesizes the final answer; (2) Planner-Coder pipeline, where a thinking model planner generates stepwise SQL generation plans and a coder synthesizes queries; and (3) Coder-Aggregator pipeline, where multiple coders independently generate SQL queries, and a reasoning agent selects the best query. Experiments on the Bird-Bench Mini-Dev set reveal that Multi-Agent discussion can improve small model performance, with up to 10.6% increase in Execution Accuracy for Qwen2.5-7b-Instruct seen after three rounds of discussion. Among the pipelines, the LLM Reasoner-Coder pipeline yields the best results, with DeepSeek-R1-32B and QwQ-32B planners boosting Gemma 3 27B IT accuracy from 52.4% to the highest score of 56.4%. Codes are available at https://github.com/treeDweller98/bappa-sql.
Authors: Asma Yamani, Malak Baslyman, Moataz Ahmed
Abstract: Large Language Models (LLMs) are increasingly employed in software engineering tasks such as requirements elicitation, design, and evaluation, raising critical questions regarding their alignment with human judgments on responsible AI values. This study investigates how closely LLMs' value preferences align with those of two human groups: a US-representative sample and AI practitioners. We evaluate 23 LLMs across four tasks: (T1) selecting key responsible AI values, (T2) rating their importance in specific contexts, (T3) resolving trade-offs between competing values, and (T4) prioritizing software requirements that embody those values. The results show that LLMs generally align more closely with AI practitioners than with the US-representative sample, emphasizing fairness, privacy, transparency, safety, and accountability. However, inconsistencies appear between the values that LLMs claim to uphold (Tasks 1-3) and the way they prioritize requirements (Task 4), revealing gaps in faithfulness between stated and applied behavior. These findings highlight the practical risk of relying on LLMs in requirements engineering without human oversight and motivate the need for systematic approaches to benchmark, interpret, and monitor value alignment in AI-assisted software development.
Authors: Fatemehzahra Darzi, Rodrigo Escobar Diaz Guerrero, Thomas Bocklitz
Abstract: Image registration refers to the process of spatially aligning two or more images by mapping them into a common coordinate system, so that corresponding anatomical or tissue structures are matched across images. In digital pathology, registration enables direct comparison and integration of information from different stains or imaging modalities, sup-porting applications such as biomarker analysis and tissue reconstruction. Accurate registration of images from different modalities is an essential step in digital pathology. In this study, we investigated how various color transformation techniques affect image registration between hematoxylin and eosin (H&E) stained images and non-linear multimodal images. We used a dataset of 20 tissue sample pairs, with each pair undergoing several preprocessing steps, including different color transformation (CycleGAN, Macenko, Reinhard, Vahadane), inversion, contrast adjustment, intensity normalization, and denoising. All images were registered using the VALIS registration method, which first applies rigid registration and then performs non-rigid registration in two steps on both low and high-resolution images. Registration performance was evaluated using the relative Target Registration Error (rTRE). We reported the median of median rTRE values (MMrTRE) and the average of median rTRE values (AMrTRE) for each method. In addition, we performed a custom point-based evaluation using ten manually selected key points. Registration was done separately for two scenarios, using either the original or inverted multimodal images. In both scenarios, CycleGAN color transformation achieved the lowest registration errors, while the other methods showed higher errors. These findings show that applying color transformation before registration improves alignment between images from different modalities and supports more reliable analysis in digital pathology.
Authors: Oshando Johnson, Alexandra Fomina, Ranjith Krishnamurthy, Vaibhav Chaudhari, Rohith Kumar Shanmuganathan, Eric Bodden
Abstract: The prevalence of security vulnerabilities has prompted companies to adopt static application security testing (SAST) tools for vulnerability detection. Nevertheless, these tools frequently exhibit usability limitations, as their generic warning messages do not sufficiently communicate important information to developers, resulting in misunderstandings or oversight of critical findings. In light of recent developments in Large Language Models (LLMs) and their text generation capabilities, our work investigates a hybrid approach that uses LLMs to tackle the SAST explainability challenges. In this paper, we present SAFE, an Integrated Development Environment (IDE) plugin that leverages GPT-4o to explain the causes, impacts, and mitigation strategies of vulnerabilities detected by SAST tools. Our expert user study findings indicate that the explanations generated by SAFE can significantly assist beginner to intermediate developers in understanding and addressing security vulnerabilities, thereby improving the overall usability of SAST tools.
Authors: Leszek Sliwko
Abstract: This paper presents a reinforced genetic approach to a defined d-resource system optimization problem. The classical evolution schema was ineffective due to a very strict feasibility function in the studied problem. Hence, the presented strategy has introduced several modifications and adaptations to standard genetic routines, e.g.: a migration operator which is an analogy to the biological random genetic drift.
Authors: Mohammed Musthafa Rafi, Adarsh Krishnamurthy, Aditya Balu
Abstract: The proliferation of AI-generated content has created an absurd communication theater where senders use LLMs to inflate simple ideas into verbose content, recipients use LLMs to compress them back into summaries, and as a consequence neither party engage with authentic content. LAAC (LLM as a Communicator) proposes a paradigm shift - positioning LLMs as intelligent communication intermediaries that capture the sender's intent through structured dialogue and facilitate genuine knowledge exchange with recipients. Rather than perpetuating cycles of AI-generated inflation and compression, LAAC enables authentic communication across diverse contexts including academic papers, proposals, professional emails, and cross-platform content generation. However, deploying LLMs as trusted communication intermediaries raises critical questions about information fidelity, consistency, and reliability. This position paper systematically evaluates the trustworthiness requirements for LAAC's deployment across multiple communication domains. We investigate three fundamental dimensions: (1) Information Capture Fidelity - accuracy of intent extraction during sender interviews across different communication types, (2) Reproducibility - consistency of structured knowledge across multiple interaction instances, and (3) Query Response Integrity - reliability of recipient-facing responses without hallucination, source conflation, or fabrication. Through controlled experiments spanning multiple LAAC use cases, we assess these trust dimensions using LAAC's multi-agent architecture. Preliminary findings reveal measurable trust gaps that must be addressed before LAAC can be reliably deployed in high-stakes communication scenarios.
Authors: Hanmo Chen, Chenghao Xu, Jiexi Yan, Cheng Deng
Abstract: Human motion style transfer allows characters to appear less rigidity and more realism with specific style. Traditional arbitrary image style transfer typically process mean and variance which is proved effective. Meanwhile, similar methods have been adapted for motion style transfer. However, due to the fundamental differences between images and motion, relying on mean and variance is insufficient to fully capture the complex dynamic patterns and spatiotemporal coherence properties of motion data. Building upon this, our key insight is to bring two more coefficient, skewness and kurtosis, into the analysis of motion style. Specifically, we propose a novel Adaptive Statistics Fusor (AStF) which consists of Style Disentanglement Module (SDM) and High-Order Multi-Statistics Attention (HOS-Attn). We trained our AStF in conjunction with a Motion Consistency Regularization (MCR) discriminator. Experimental results show that, by providing a more comprehensive model of the spatiotemporal statistical patterns inherent in dynamic styles, our proposed AStF shows proficiency superiority in motion style transfers over state-of-the-arts. Our code and model are available at https://github.com/CHMimilanlan/AStF.
Authors: Hikari Otsuka, Daiki Chijiwa, Yasuyuki Okoshi, Daichi Fujiki, Susumu Takeuchi, Masato Motomura
Abstract: The strong lottery ticket hypothesis (SLTH) conjectures that high-performing subnetworks, called strong lottery tickets (SLTs), are hidden in randomly initialized neural networks. Although recent theoretical studies have established the SLTH across various neural architectures, the SLTH for transformer architectures still lacks theoretical understanding. In particular, the current theory of the SLTH does not yet account for the multi-head attention (MHA) mechanism, a core component of transformers. To address this gap, we introduce a theoretical analysis of the existence of SLTs within MHAs. We prove that, if a randomly initialized MHA of $H$ heads and input dimension $d$ has the hidden dimension $O(d\log(Hd^{3/2}))$ for the key and value, it contains an SLT that approximates an arbitrary MHA with the same input dimension with high probability. Furthermore, by leveraging this theory for MHAs, we extend the SLTH to transformers without normalization layers. We empirically validate our theoretical findings, demonstrating that the approximation error between the SLT within a source model (MHA and transformer) and an approximate target counterpart decreases exponentially by increasing the hidden dimension of the source model.
Authors: Zefeng Li, Ning Yang
Abstract: Although the collaborative filtering (CF) algorithm has achieved remarkable performance in recommendation systems, it suffers from suboptimal recommendation performance due to noise in the user-item interaction matrix. Numerous noise-removal studies have improved recommendation models, but most existing approaches conduct denoising on a single graph. This may cause attenuation of collaborative signals: removing edges between two nodes can interrupt paths between other nodes, weakening path-dependent collaborative information. To address these limitations, this study proposes a novel GNN-based CF model called DRCSD for denoising unstable interactions. DRCSD includes two core modules: a collaborative signal decoupling module (decomposes signals into distinct orders by structural characteristics) and an order-wise denoising module (performs targeted denoising on each order). Additionally, the information aggregation mechanism of traditional GNN-based CF models is modified to avoid cross-order signal interference until the final pooling operation. Extensive experiments on three public real-world datasets show that DRCSD has superior robustness against unstable interactions and achieves statistically significant performance improvements in recommendation accuracy metrics compared to state-of-the-art baseline models.
Authors: Rasmus M{\o}ller-Larsen, Adam Izdebski, Jan Olszewski, Pankhil Gawade, Michal Kmicikiewicz, Wojciech Zarzecki, Ewa Szczurek
Abstract: Recent advances in computational methods for designing biological sequences have sparked the development of metrics to evaluate these methods performance in terms of the fidelity of the designed sequences to a target distribution and their attainment of desired properties. However, a single software library implementing these metrics was lacking. In this work we introduce seqme, a modular and highly extendable open-source Python library, containing model-agnostic metrics for evaluating computational methods for biological sequence design. seqme considers three groups of metrics: sequence-based, embedding-based, and property-based, and is applicable to a wide range of biological sequences: small molecules, DNA, ncRNA, mRNA, peptides and proteins. The library offers a number of embedding and property models for biological sequences, as well as diagnostics and visualization functions to inspect the results. seqme can be used to evaluate both one-shot and iterative computational design methods.
Authors: Allie Tran, Luca Rossetto
Abstract: Multimodal co-embedding models, especially CLIP, have advanced the state of the art in zero-shot classification and multimedia information retrieval in recent years by aligning images and text in a shared representation space. However, such modals trained on a contrastive alignment can lack stability towards small input perturbations. Especially when dealing with manually expressed queries, minor variations in the query can cause large differences in the ranking of the best-matching results. In this paper, we present a systematic analysis of the effect of multiple classes of non-semantic query perturbations in an multimedia information retrieval scenario. We evaluate a diverse set of lexical, syntactic, and semantic perturbations across multiple CLIP variants using the TRECVID Ad-Hoc Video Search queries and the V3C1 video collection. Across models, we find that syntactic and semantic perturbations drive the largest instabilities, while brittleness is concentrated in trivial surface edits such as punctuation and case. Our results highlight robustness as a critical dimension for evaluating vision-language models beyond benchmark accuracy.
Authors: Marawan Elbatel, Anbang Wang, Keyuan Liu, Kaouther Mouheb, Enrique Almar-Munoz, Lizhuo Lin, Yanqi Yang, Karim Lekadir, Xiaomeng Li
Abstract: This paper does not introduce a novel architecture; instead, it revisits a fundamental yet overlooked baseline: adapting human-centric foundation models for anatomical landmark detection in medical imaging. While landmark detection has traditionally relied on domain-specific models, the emergence of large-scale pre-trained vision models presents new opportunities. In this study, we investigate the adaptation of Sapiens, a human-centric foundation model designed for pose estimation, to medical imaging through multi-dataset pretraining, establishing a new state of the art across multiple datasets. Our proposed model, MedSapiens, demonstrates that human-centric foundation models, inherently optimized for spatial pose localization, provide strong priors for anatomical landmark detection, yet this potential has remained largely untapped. We benchmark MedSapiens against existing state-of-the-art models, achieving up to 5.26% improvement over generalist models and up to 21.81% improvement over specialist models in the average success detection rate (SDR). To further assess MedSapiens adaptability to novel downstream tasks with few annotations, we evaluate its performance in limited-data settings, achieving 2.69% improvement over the few-shot state of the art in SDR. Code and model weights are available at https://github.com/xmed-lab/MedSapiens .
Authors: Claudio Giusti, Luca Guarnera, Sebastiano Battiato
Abstract: The growing sophistication of synthetic image and deepfake generation models has turned source attribution and authenticity verification into a critical challenge for modern computer vision systems. Recent studies suggest that diffusion pipelines unintentionally imprint persistent statistical traces, known as signal leaks, within their outputs, particularly in latent representations. Building on this observation, we propose Proto-LeakNet, a signal-leak-aware and interpretable attribution framework that integrates closed-set classification with a density-based open-set evaluation on the learned embeddings, enabling analysis of unseen generators without retraining. Operating in the latent domain of diffusion models, our method re-simulates partial forward diffusion to expose residual generator-specific cues. A temporal attention encoder aggregates multi-step latent features, while a feature-weighted prototype head structures the embedding space and enables transparent attribution. Trained solely on closed data and achieving a Macro AUC of 98.13%, Proto-LeakNet learns a latent geometry that remains robust under post-processing, surpassing state-of-the-art methods, and achieves strong separability between known and unseen generators. These results demonstrate that modeling signal-leak bias in latent space enables reliable and interpretable AI-image and deepfake forensics. The code for the whole work will be available upon submission.
Authors: Matteo Cercola, Valeria Capretti, Simone Formentin
Abstract: Learning from human preferences is a cornerstone of aligning machine learning models with subjective human judgments. Yet, collecting such preference data is often costly and time-consuming, motivating the need for more efficient learning paradigms. Two established approaches offer complementary advantages: RLHF scales effectively to high-dimensional tasks such as LLM fine-tuning, while PBO achieves greater sample efficiency through active querying. We propose a hybrid framework that unifies RLHF's scalability with PBO's query efficiency by integrating an acquisition-driven module into the RLHF pipeline, thereby enabling active and sample-efficient preference gathering. We validate the proposed approach on two representative domains: (i) high-dimensional preference optimization and (ii) LLM fine-tuning. Experimental results demonstrate consistent improvements in both sample efficiency and overall performance across these tasks.
Authors: Robin Spanier, Thorsten Hoeser, Claudia Kuenzer
Abstract: The recent and ongoing expansion of marine infrastructure, including offshore wind farms, oil and gas platforms, artificial islands, and aquaculture facilities, highlights the need for effective monitoring systems. The development of robust models for offshore infrastructure detection relies on comprehensive, balanced datasets, but falls short when samples are scarce, particularly for underrepresented object classes, shapes, and sizes. By training deep learning-based YOLOv10 object detection models with a combination of synthetic and real Sentinel-1 satellite imagery acquired in the fourth quarter of 2023 from four regions (Caspian Sea, South China Sea, Gulf of Guinea, and Coast of Brazil), this study investigates the use of synthetic training data to enhance model performance. We evaluated this approach by applying the model to detect offshore platforms in three unseen regions (Gulf of Mexico, North Sea, Persian Gulf) and thereby assess geographic transferability. This region-holdout evaluation demonstrated that the model generalises beyond the training areas. In total, 3,529 offshore platforms were detected, including 411 in the North Sea, 1,519 in the Gulf of Mexico, and 1,593 in the Persian Gulf. The model achieved an F1 score of 0.85, which improved to 0.90 upon incorporating synthetic data. We analysed how synthetic data enhances the representation of unbalanced classes and overall model performance, taking a first step toward globally transferable detection of offshore infrastructure. This study underscores the importance of balanced datasets and highlights synthetic data generation as an effective strategy to address common challenges in remote sensing, demonstrating the potential of deep learning for scalable, global offshore infrastructure monitoring.
Authors: Yuanpeng Zhang, Xing Hu, Xi Chen, Zhihang Yuan, Cong Li, Jingchen Zhu, Zhao Wang, Chenguang Zhang, Xin Si, Wei Gao, Qiang Wu, Runsheng Wang, Guangyu Sun
Abstract: SRAM Processing-in-Memory (PIM) has emerged as the most promising implementation for high-performance PIM, delivering superior computing density, energy efficiency, and computational precision. However, the pursuit of higher performance necessitates more complex circuit designs and increased operating frequencies, which exacerbate IR-drop issues. Severe IR-drop can significantly degrade chip performance and even threaten reliability. Conventional circuit-level IR-drop mitigation methods, such as back-end optimizations, are resource-intensive and often compromise power, performance, and area (PPA). To address these challenges, we propose AIM, comprehensive software and hardware co-design for architecture-level IR-drop mitigation in high-performance PIM. Initially, leveraging the bit-serial and in-situ dataflow processing properties of PIM, we introduce Rtog and HR, which establish a direct correlation between PIM workloads and IR-drop. Building on this foundation, we propose LHR and WDS, enabling extensive exploration of architecture-level IR-drop mitigation while maintaining computational accuracy through software optimization. Subsequently, we develop IR-Booster, a dynamic adjustment mechanism that integrates software-level HR information with hardware-based IR-drop monitoring to adapt the V-f pairs of the PIM macro, achieving enhanced energy efficiency and performance. Finally, we propose the HR-aware task mapping method, bridging software and hardware designs to achieve optimal improvement. Post-layout simulation results on a 7nm 256-TOPS PIM chip demonstrate that AIM achieves up to 69.2% IR-drop mitigation, resulting in 2.29x energy efficiency improvement and 1.152x speedup.
Authors: Antti Koskela, Tejas Kulkarni, Laith Zumot
Abstract: Differentially private in-context learning (DP-ICL) has recently become an active research topic due to the inherent privacy risks of in-context learning. However, existing approaches overlook a critical component of modern large language model (LLM) pipelines: the similarity search used to retrieve relevant context data. In this work, we introduce a DP framework for in-context learning that integrates nearest neighbor search of relevant examples in a privacy-aware manner. Our method outperforms existing baselines by a substantial margin across all evaluated benchmarks, achieving more favorable privacy-utility trade-offs. To achieve this, we employ nearest neighbor retrieval from a database of context data, combined with a privacy filter that tracks the cumulative privacy cost of selected samples to ensure adherence to a central differential privacy budget. Experimental results on text classification and document question answering show a clear advantage of the proposed method over existing baselines.
Authors: Federico Pirola, Fabio Stella, Marco Grzegorczyk
Abstract: Dynamic Bayesian networks (DBNs) are increasingly used in healthcare due to their ability to model complex temporal relationships in patient data while maintaining interpretability, an essential feature for clinical decision-making. However, existing approaches to handling missing data in longitudinal clinical datasets are largely derived from static Bayesian networks literature, failing to properly account for the temporal nature of the data. This gap limits the ability to quantify uncertainty over time, which is particularly critical in settings such as intensive care, where understanding the temporal dynamics is fundamental for model trustworthiness and applicability across diverse patient groups. Despite the potential of DBNs, a full Bayesian framework that integrates missing data handling remains underdeveloped. In this work, we propose a novel Gibbs sampling-based method for learning DBNs from incomplete data. Our method treats each missing value as an unknown parameter following a Gaussian distribution. At each iteration, the unobserved values are sampled from their full conditional distributions, allowing for principled imputation and uncertainty estimation. We evaluate our method on both simulated datasets and real-world intensive care data from critically ill patients. Compared to standard model-agnostic techniques such as MICE, our Bayesian approach demonstrates superior reconstruction accuracy and convergence properties. These results highlight the clinical relevance of incorporating full Bayesian inference in temporal models, providing more reliable imputations and offering deeper insight into model behavior. Our approach supports safer and more informed clinical decision-making, particularly in settings where missing data are frequent and potentially impactful.
Authors: Ali Boudaghi, Hadi Zare
Abstract: Music editing has emerged as an important and practical area of artificial intelligence, with applications ranging from video game and film music production to personalizing existing tracks according to user preferences. However, existing models face significant limitations, such as being restricted to editing synthesized music generated by their own models, requiring highly precise prompts, or necessitating task-specific retraining, thus lacking true zero-shot capability. Leveraging recent advances in rectified flow and diffusion transformers, we introduce MusRec, the first zero-shot text-to-music editing model capable of performing diverse editing tasks on real-world music efficiently and effectively. Experimental results demonstrate that our approach outperforms existing methods in preserving musical content, structural consistency, and editing fidelity, establishing a strong foundation for controllable music editing in real-world scenarios.
Authors: Subeen Park, Joowang Kim, Hakyung Lee, Sunjae Yoo, Kyungwoo Song
Abstract: Deep learning models achieve strong performance across various domains but often rely on spurious correlations, making them vulnerable to distribution shifts. This issue is particularly severe in subpopulation shift scenarios, where models struggle in underrepresented groups. While existing methods have made progress in mitigating this issue, their performance gains are still constrained. They lack a rigorous theoretical framework connecting the embedding space representations with worst-group error. To address this limitation, we propose Spurious Correlation-Aware Embedding Regularization for Worst-Group Robustness (SCER), a novel approach that directly regularizes feature representations to suppress spurious cues. We show theoretically that worst-group error is influenced by how strongly the classifier relies on spurious versus core directions, identified from differences in group-wise mean embeddings across domains and classes. By imposing theoretical constraints at the embedding level, SCER encourages models to focus on core features while reducing sensitivity to spurious patterns. Through systematic evaluation on multiple vision and language, we show that SCER outperforms prior state-of-the-art studies in worst-group accuracy. Our code is available at \href{https://github.com/MLAI-Yonsei/SCER}{https://github.com/MLAI-Yonsei/SCER}.
URLs: https://github.com/MLAI-Yonsei/SCER, https://github.com/MLAI-Yonsei/SCER
Authors: Jayadeva, Naman Dwivedi, Hari Krishnan, N. M. Anoop Krishnan
Abstract: A formal link between regression and classification has been tenuous. Even though the margin maximization term $\|w\|$ is used in support vector regression, it has at best been justified as a regularizer. We show that a regression problem with $M$ samples lying on a hyperplane has a one-to-one equivalence with a linearly separable classification task with $2M$ samples. We show that margin maximization on the equivalent classification task leads to a different regression formulation than traditionally used. Using the equivalence, we demonstrate a ``regressability'' measure, that can be used to estimate the difficulty of regressing a dataset, without needing to first learn a model for it. We use the equivalence to train neural networks to learn a linearizing map, that transforms input variables into a space where a linear regressor is adequate.
Authors: Hao He, Courtney Miller, Shyam Agarwal, Christian K\"astner, Bogdan Vasilescu
Abstract: Large language models (LLMs) have demonstrated the promise to revolutionize the field of software engineering. Among other things, LLM agents are rapidly gaining momentum in their application to software development, with practitioners claiming a multifold productivity increase after adoption. Yet, empirical evidence is lacking around these claims. In this paper, we estimate the causal effect of adopting a widely popular LLM agent assistant, namely Cursor, on development velocity and software quality. The estimation is enabled by a state-of-the-art difference-in-differences design comparing Cursor-adopting GitHub projects with a matched control group of similar GitHub projects that do not use Cursor. We find that the adoption of Cursor leads to a significant, large, but transient increase in project-level development velocity, along with a significant and persistent increase in static analysis warnings and code complexity. Further panel generalized method of moments estimation reveals that the increase in static analysis warnings and code complexity acts as a major factor causing long-term velocity slowdown. Our study carries implications for software engineering practitioners, LLM agent assistant designers, and researchers.
Authors: Patrik Val\'abek, Michaela Horv\'athov\'a, Martin Klau\v{c}o
Abstract: This paper presents a deep Koopman-based Economic Model Predictive Control (EMPC) for efficient operation of a laboratory-scale pasteurization unit (PU). The method uses Koopman operator theory to transform the complex, nonlinear system dynamics into a linear representation, enabling the application of convex optimization while representing the complex PU accurately. The deep Koopman model utilizes neural networks to learn the linear dynamics from experimental data, achieving a 45% improvement in open-loop prediction accuracy over conventional N4SID subspace identification. Both analyzed models were employed in the EMPC formulation that includes interpretable economic costs, such as energy consumption, material losses due to inadequate pasteurization, and actuator wear. The feasibility of EMPC is ensured using slack variables. The deep Koopman EMPC and N4SID EMPC are numerically validated on a nonlinear model of multivariable PU under external disturbance. The disturbances include feed pump fail-to-close scenario and the introduction of a cold batch to be pastuerized. These results demonstrate that the deep Koopmand EMPC achieves a 32% reduction in total economic cost compared to the N4SID baseline. This improvement is mainly due to the reductions in material losses and energy consumption. Furthermore, the steady-state operation via Koopman-based EMPC requires 10.2% less electrical energy. The results highlight the practical advantages of integrating deep Koopman representations with economic optimization to achieve resource-efficient control of thermal-intensive plants.
Authors: Patrik Val\'abek, Marek Wadinger, Michal Kvasnica, Martin Klau\v{c}o
Abstract: Nonlinear dynamical systems with input delays pose significant challenges for prediction, estimation, and control due to their inherent complexity and the impact of delays on system behavior. Traditional linear control techniques often fail in these contexts, necessitating innovative approaches. This paper introduces a novel approach to approximate the Koopman operator using an LSTM-enhanced Deep Koopman model, enabling linear representations of nonlinear systems with time delays. By incorporating Long Short-Term Memory (LSTM) layers, the proposed framework captures historical dependencies and efficiently encodes time-delayed system dynamics into a latent space. Unlike traditional extended Dynamic Mode Decomposition (eDMD) approaches that rely on predefined dictionaries, the LSTM-enhanced Deep Koopman model is dictionary-free, which mitigates the problems with the underlying dynamics being known and incorporated into the dictionary. Quantitative comparisons with extended eDMD on a simulated system demonstrate highly significant performance gains in prediction accuracy in cases where the true nonlinear dynamics are unknown and achieve comparable results to eDMD with known dynamics of a system.
Authors: Abheek Ghosh, Tzeh Yuan Neoh, Nicholas Teh, Giannis Tyrovolas
Abstract: We study a model of subscription-based platforms where users pay a fixed fee for unlimited access to content, and creators receive a share of the revenue. Existing approaches to detecting fraud predominantly rely on machine learning methods, engaging in an ongoing arms race with bad actors. We explore revenue division mechanisms that inherently disincentivize manipulation. We formalize three types of manipulation-resistance axioms and examine which existing rules satisfy these. We show that a mechanism widely used by streaming platforms, not only fails to prevent fraud, but also makes detecting manipulation computationally intractable. We also introduce a novel rule, ScaledUserProp, that satisfies all three manipulation-resistance axioms. Finally, experiments with both real-world and synthetic streaming data support ScaledUserProp as a fairer alternative compared to existing rules.
Authors: Alberto Cattaneo, Carlo Luschi, Daniel Justus
Abstract: Retrieval of information from graph-structured knowledge bases represents a promising direction for improving the factuality of LLMs. While various solutions have been proposed, a comparison of methods is difficult due to the lack of challenging QA datasets with ground-truth targets for graph retrieval. We present SynthKGQA, a framework for generating high-quality synthetic Knowledge Graph Question Answering datasets from any Knowledge Graph, providing the full set of ground-truth facts in the KG to reason over each question. We show how, in addition to enabling more informative benchmarking of KG retrievers, the data produced with SynthKGQA also allows us to train better models. We apply SynthKGQA to Wikidata to generate GTSQA, a new dataset designed to test zero-shot generalization abilities of KG retrievers with respect to unseen graph structures and relation types, and benchmark popular solutions for KG-augmented LLMs on it.
Authors: Hyo Jin Do, Zahra Ashktorab, Jasmina Gajcin, Erik Miehling, Mart\'in Santill\'an Cooper, Qian Pan, Elizabeth M. Daly, Werner Geyer
Abstract: The LLM-as-a-judge paradigm enables flexible, user-defined evaluation, but its effectiveness is often limited by the scarcity of diverse, representative data for refining criteria. We present a tool that integrates synthetic data generation into the LLM-as-a-judge workflow, empowering users to create tailored and challenging test cases with configurable domains, personas, lengths, and desired outcomes, including borderline cases. The tool also supports AI-assisted inline editing of existing test cases. To enhance transparency and interpretability, it reveals the prompts and explanations behind each generation. In a user study (N=24), 83% of participants preferred the tool over manually creating or selecting test cases, as it allowed them to rapidly generate diverse synthetic data without additional workload. The generated synthetic data proved as effective as hand-crafted data for both refining evaluation criteria and aligning with human preferences. These findings highlight synthetic data as a promising alternative, particularly in contexts where efficiency and scalability are critical.
Authors: Ipsita Ghosh, Ethan Nguyen, Christian K\"ummerle
Abstract: Parameter-efficient training, based on low-rank optimization, has become a highly successful tool for fine-tuning large deep-learning models. However, these methods fail at low-rank pre-training tasks where maintaining the low-rank structure and the objective remains a challenging task. We propose the Quadratic Reweighted Rank Regularizer dubbed Q3R, which leads to a novel low-rank inducing training strategy inspired by the iteratively reweighted least squares (IRLS) framework. Q3R is based on a quadratic regularizer term which majorizes a smoothed log determinant serving as rank surrogate objective. Unlike other low-rank training techniques, Q3R is able to train weight matrices with prescribed, low target ranks of models that achieve comparable predictive performance as dense models, with small computational overhead, while remaining fully compatible with existing architectures. For example, we demonstrated one experiment where we are able to truncate $60\%$ and $80\%$ of the parameters of a ViT-Tiny model with $~1.3\%$ and $~4\%$ accuracy drop in CIFAR-10 performance respectively. The efficacy of Q3R is confirmed on Transformers across both image and language tasks, including for low-rank fine-tuning.
Authors: Nikhil Abhyankar, Purvi Chaurasia, Sanchit Kabra, Ananya Srivastava, Vivek Gupta, Chandan K. Reddy
Abstract: Existing tabular reasoning benchmarks mostly test models on small, uniform tables, underrepresenting the complexity of real-world data and giving an incomplete view of Large Language Models' (LLMs) reasoning abilities. Real tables are long, heterogeneous, and domain-specific, mixing structured fields with free text and requiring multi-hop reasoning across thousands of tokens. To address this gap, we introduce RUST-BENCH, a benchmark of 7966 questions from 2031 real-world tables spanning two domains: i) RB-Science (NSF grant records) and ii) RB-Sports (NBA statistics). Unlike prior work, RUST-BENCH evaluates LLMs jointly across scale, heterogeneity, domain specificity, and reasoning complexity. Experiments with open-source and proprietary models show that LLMs struggle with heterogeneous schemas and complex multi-hop inference, revealing persistent weaknesses in current architectures and prompting strategies. RUST-BENCH establishes a challenging new testbed for advancing tabular reasoning research.
Authors: Cuong Huynh, Jie Cao
Abstract: This paper describes the OUNLP system submitted to the TSAR-2025 Shared Task (Alva-Manchego et al., 2025), designed for readability-controlled text simplification using LLM-prompting-based generation. Based on the analysis of prompt-based text simplification methods, we discovered an interesting finding that text simplification performance is highly related to the gap between the source CEFR (Arase et al., 2022) level and the target CEFR level. Inspired by this finding, we propose two multi-round simplification methods and generate them via GPT-4o: rule-based simplification (MRS-Rule) and jointly rule-based LLM simplification (MRS-Joint). Our submitted systems ranked 7 out of 20 teams. Later improvements with MRS-Joint show that taking the LLM simplified candidates as the starting point could further boost the multi-round simplification performance.
Authors: Christos-Nikolaos Zacharopoulos, Revekka Kyriakoglou
Abstract: As Large Language Models (LLMs) become integral to human-centered applications, understanding their personality-like behaviors is increasingly important for responsible development and deployment. This paper systematically evaluates six LLMs, applying the Big Five Inventory-2 (BFI-2) framework, to assess trait expressions under varying sampling temperatures. We find significant differences across four of the five personality dimensions, with Neuroticism and Extraversion susceptible to temperature adjustments. Further, hierarchical clustering reveals distinct model clusters, suggesting that architectural features may predispose certain models toward stable trait profiles. Taken together, these results offer new insights into the emergence of personality-like patterns in LLMs and provide a new perspective on model tuning, selection, and the ethical governance of AI systems. We share the data and code for this analysis here: https://osf.io/bsvzc/?view_only=6672219bede24b4e875097426dc3fac1
URLs: https://osf.io/bsvzc/?view_only=6672219bede24b4e875097426dc3fac1
Authors: Joshua Gao, Quoc Huy Pham, Subin Varghese, Silwal Saurav, Vedhus Hoskere
Abstract: Retrieval-Augmented Generation (RAG) is a critical technique for grounding Large Language Models (LLMs) in factual evidence, yet evaluating RAG systems in specialized, safety-critical domains remains a significant challenge. Existing evaluation frameworks often rely on heuristic-based metrics that fail to capture domain-specific nuances and other works utilize LLM-as-a-Judge approaches that lack validated alignment with human judgment. This paper introduces RAGalyst, an automated, human-aligned agentic framework designed for the rigorous evaluation of domain-specific RAG systems. RAGalyst features an agentic pipeline that generates high-quality, synthetic question-answering (QA) datasets from source documents, incorporating an agentic filtering step to ensure data fidelity. The framework refines two key LLM-as-a-Judge metrics-Answer Correctness and Answerability-using prompt optimization to achieve a strong correlation with human annotations. Applying this framework to evaluate various RAG components across three distinct domains (military operations, cybersecurity, and bridge engineering), we find that performance is highly context-dependent. No single embedding model, LLM, or hyperparameter configuration proves universally optimal. Additionally, we provide an analysis on the most common low Answer Correctness reasons in RAG. These findings highlight the necessity of a systematic evaluation framework like RAGalyst, which empowers practitioners to uncover domain-specific trade-offs and make informed design choices for building reliable and effective RAG systems. RAGalyst is available on our Github.
Authors: Shaolong Wu, James Blume, Geshi Yeung
Abstract: Algorithmic fairness has grown rapidly as a research area, yet key concepts remain unsettled, especially in criminal justice. We review group, individual, and process fairness and map the conditions under which they conflict. We then develop a simple modification to standard group fairness. Rather than exact parity across protected groups, we minimize a weighted error loss while keeping differences in false negative rates within a small tolerance. This makes solutions easier to find, can raise predictive accuracy, and surfaces the ethical choice of error costs. We situate this proposal within three classes of critique: biased and incomplete data, latent affirmative action, and the explosion of subgroup constraints. Finally, we offer a practical framework for deployment in public decision systems built on three pillars: need-based decisions, Transparency and accountability, and narrowly tailored definitions and solutions. Together, these elements link technical design to legitimacy and provide actionable guidance for agencies that use risk assessment and related tools.
Authors: Amir Zur, Atticus Geiger, Ekdeep Singh Lubana, Eric Bigelow
Abstract: When a language model generates text, the selection of individual tokens might lead it down very different reasoning paths, making uncertainty difficult to quantify. In this work, we consider whether reasoning language models represent the alternate paths that they could take during generation. To test this hypothesis, we use hidden activations to control and predict a language model's uncertainty during chain-of-thought reasoning. In our experiments, we find a clear correlation between how uncertain a model is at different tokens, and how easily the model can be steered by controlling its activations. This suggests that activation interventions are most effective when there are alternate paths available to the model -- in other words, when it has not yet committed to a particular final answer. We also find that hidden activations can predict a model's future outcome distribution, demonstrating that models implicitly represent the space of possible paths.
Authors: Baptiste Bonin, Maxime Heuillet, Audrey Durand
Abstract: Modeling user preferences across domains remains a key challenge in slate recommendation (i.e. recommending an ordered sequence of items) research. We investigate how Large Language Models (LLM) can effectively act as world models of user preferences through pairwise reasoning over slates. We conduct an empirical study involving several LLMs on three tasks spanning different datasets. Our results reveal relationships between task performance and properties of the preference function captured by LLMs, hinting towards areas for improvement and highlighting the potential of LLMs as world models in recommender systems.
Authors: Divyansha Lachi, Mahmoud Mohammadi, Joe Meyer, Vinam Arora, Tom Palczewski, Eva L. Dyer
Abstract: In domains such as healthcare, finance, and e-commerce, the temporal dynamics of relational data emerge from complex interactions-such as those between patients and providers, or users and products across diverse categories. To be broadly useful, models operating on these data must integrate long-range spatial and temporal dependencies across diverse types of entities, while also supporting multiple predictive tasks. However, existing graph models for relational data primarily focus on spatial structure, treating temporal information merely as a filtering constraint to exclude future events rather than a modeling signal, and are typically designed for single-task prediction. To address these gaps, we introduce a temporal subgraph sampler that enhances global context by retrieving nodes beyond the immediate neighborhood to capture temporally relevant relationships. In addition, we propose the Relational Graph Perceiver (RGP), a graph transformer architecture for relational deep learning that leverages a cross-attention-based latent bottleneck to efficiently integrate information from both structural and temporal contexts. This latent bottleneck integrates signals from different node and edge types into a common latent space, enabling the model to build global context across the entire relational system. RGP also incorporates a flexible cross-attention decoder that supports joint learning across tasks with disjoint label spaces within a single model. Experiments on RelBench, SALT, and CTU show that RGP delivers state-of-the-art performance, offering a general and scalable solution for relational deep learning with support for diverse predictive tasks.
Authors: Satchel Grant, Simon Jerome Han, Alexa Tartaglini, Christopher Potts
Abstract: A common approach to mechanistic interpretability is to causally manipulate model representations via targeted interventions in order to understand what those representations encode. Here we ask whether such interventions create out-of-distribution (divergent) representations, and whether this raises concerns about how faithful their resulting explanations are to the target model in its natural state. First, we demonstrate empirically that common causal intervention techniques often do shift internal representations away from the natural distribution of the target model. Then, we provide a theoretical analysis of two classes of such divergences: `harmless' divergences that occur in the null-space of the weights and from covariance within behavioral decision boundaries, and `pernicious' divergences that activate hidden network pathways and cause dormant behavioral changes. Finally, in an effort to mitigate the pernicious cases, we modify the Counterfactual Latent (CL) loss from Grant (2025) that regularizes interventions to remain closer to the natural distributions, reducing the likelihood of harmful divergences while preserving the interpretive power of interventions. Together, these results highlight a path towards more reliable interpretability methods.
Authors: Maximus A. Pace, Prithwish Dan, Chuanruo Ning, Atiksh Bhardwaj, Audrey Du, Edward W. Duan, Wei-Chiu Ma, Kushal Kedia
Abstract: Human videos can be recorded quickly and at scale, making them an appealing source of training data for robot learning. However, humans and robots differ fundamentally in embodiment, resulting in mismatched action execution. Direct kinematic retargeting of human hand motion can therefore produce actions that are physically infeasible for robots. Despite these low-level differences, human demonstrations provide valuable motion cues about how to manipulate and interact with objects. Our key idea is to exploit the forward diffusion process: as noise is added to actions, low-level execution differences fade while high-level task guidance is preserved. We present X-Diffusion, a principled framework for training diffusion policies that maximally leverages human data without learning dynamically infeasible motions. X-Diffusion first trains a classifier to predict whether a noisy action is executed by a human or robot. Then, a human action is incorporated into policy training only after adding sufficient noise such that the classifier cannot discern its embodiment. Actions consistent with robot execution supervise fine-grained denoising at low noise levels, while mismatched human actions provide only coarse guidance at higher noise levels. Our experiments show that naive co-training under execution mismatches degrades policy performance, while X-Diffusion consistently improves it. Across five manipulation tasks, X-Diffusion achieves a 16% higher average success rate than the best baseline. The project website is available at https://portal-cornell.github.io/X-Diffusion/.
Authors: Ryuta Arisaka
Abstract: We make three contributions. First, we formulate a discussion-graph semantics for first-order logic with equality, enabling reasoning about discussion and argumentation in AI more generally than before. This addresses the current lack of a formal reasoning framework capable of handling diverse discussion and argumentation models. Second, we generalise Dung's notion of extensions to cases where two or more graph nodes in an argumentation framework are equivalent. Third, we connect these two contributions by showing that the generalised extensions are first-order characterisable within the proposed discussion-graph semantics. Propositional characterisability of all Dung's extensions is an immediate consequence.
Authors: Chuan Tian, Yilei Zhang
Abstract: Large Language Model (LLM)-based multi-agent systems are increasingly applied to automate computational workflows in science and engineering. However, how inter-agent dynamics influence reasoning quality and verification reliability remains unclear. We study these mechanisms using an AutoGen-based multi-agent framework for linear-elastic Finite Element Analysis (FEA), evaluating seven role configurations across four tasks under a fixed 12-turn conversation limit. From 1,120 controlled trials, we find that collaboration effectiveness depends more on functional complementarity than team size: the three-agent Coder-Executor-Critic configuration uniquely produced physically and visually correct solutions, while adding redundant reviewers reduced success rates. Yet three systematic failure modes persist: (1) affirmation bias, where the Rebuttal agent endorsed rather than challenged outputs (85-92% agreement, including errors); (2) premature consensus caused by redundant reviewers; and (3) a verification-validation gap where executable but physically incorrect code passed undetected. No agent combination successfully validated constitutive relations in complex tasks. Building on theories of functional diversity, role differentiation, and computational validation, we propose actionable design principles: (i) assign complementary agent roles, (ii) enforce multi-level validation (execution, specification, physics), and (iii) prevent early consensus through adversarial or trigger-based interaction control. These findings establish a principled foundation for designing trustworthy LLM collaborations in engineering workflows.
Authors: Seema Aswani, Sujala D. Shetty
Abstract: Explainable Artificial Intelligence (XAI) methods in text summarization are essential for understanding the model behavior and fostering trust in model-generated summaries. Despite the effectiveness of XAI methods, recent studies have highlighted a key challenge in this area known as the "disagreement problem". This problem occurs when different XAI methods yield conflicting explanations for the same model outcome. Such discrepancies raise concerns about the consistency of explanations and reduce confidence in model interpretations, which is crucial for secure and accountable AI applications. This work is among the first to empirically investigate the disagreement problem in text summarization, demonstrating that such discrepancies are widespread in state-of-the-art summarization models. To address this gap, we propose Regional Explainable AI (RXAI) a novel segmentation-based approach, where each article is divided into smaller, coherent segments using sentence transformers and clustering. We use XAI methods on text segments to create localized explanations that help reduce disagreement between different XAI methods, thereby enhancing the trustworthiness of AI-generated summaries. Our results illustrate that the localized explanations are more consistent than full-text explanations. The proposed approach is validated using two benchmark summarization datasets, Extreme summarization (Xsum) and CNN/Daily Mail, indicating a substantial decrease in disagreement. Additionally, the interactive JavaScript visualization tool is developed to facilitate easy, color-coded exploration of attribution scores at the sentence level, enhancing user comprehension of model explanations.
Authors: Feifei Zhao, Hui Feng, Haibo Tong, Zhengqiang Han, Erliang Lin, Enmeng Lu, Yinqian Sun, Yi Zeng
Abstract: As AI closely interacts with human society, it is crucial to ensure that its behavior is safe, altruistic, and aligned with human ethical and moral values. However, existing research on embedding ethical considerations into AI remains insufficient, and previous external constraints based on principles and rules are inadequate to provide AI with long-term stability and generalization capabilities. Emotional empathy intrinsically motivates altruistic behaviors aimed at alleviating others' negative emotions through emotional sharing and contagion mechanisms. Motivated by this, we draw inspiration from the neural mechanism of human emotional empathy-driven altruistic decision making, and simulate the shared self-other perception-mirroring-empathy neural circuits, to construct a brain-inspired emotional empathy-driven altruistic decision-making model. Here, empathy directly impacts dopamine release to form intrinsic altruistic motivation. The proposed model exhibits consistent altruistic behaviors across three experimental settings: emotional contagion-integrated two-agent altruistic rescue, multi-agent gaming, and robotic emotional empathy interaction scenarios. In-depth analyses validate the positive correlation between empathy levels and altruistic preferences (consistent with psychological behavioral experiment findings), while also demonstrating how interaction partners' empathy levels influence the agent's behavioral patterns. We further test the proposed model's performance and stability in moral dilemmas involving conflicts between self-interest and others' well-being, partially observable environments, and adversarial defense scenarios. This work provides preliminary exploration of human-like empathy-driven altruistic moral decision making, contributing potential perspectives for developing ethically-aligned AI.
Authors: Dutao Zhang, Nicolas Rafael Arroyo Arias, YuLong He, Sergey Kovalchuk
Abstract: Controllable code generation, the ability to synthesize code that follows a specified style while maintaining functionality, remains a challenging task. We propose a two-stage training framework combining contrastive learning and conditional decoding to enable flexible style control. The first stage aligns code style representations with semantic and structural features. In the second stage, we fine-tune a language model (e.g., Flan-T5) conditioned on the learned style vector to guide generation. Our method supports style interpolation and user personalization via lightweight mixing. Compared to prior work, our unified framework offers improved stylistic control without sacrificing code correctness. This is among the first approaches to combine contrastive alignment with conditional decoding for style-guided code generation.
Authors: Yichi Zhang, Jinlong Pang, Zhaowei Zhu, Yang Liu
Abstract: The recent success of generative AI highlights the crucial role of high-quality human feedback in building trustworthy AI systems. However, the increasing use of large language models (LLMs) by crowdsourcing workers poses a significant challenge: datasets intended to reflect human input may be compromised by LLM-generated responses. Existing LLM detection approaches often rely on high-dimensional training data such as text, making them unsuitable for annotation tasks like multiple-choice labeling. In this work, we investigate the potential of peer prediction -- a mechanism that evaluates the information within workers' responses without using ground truth -- to mitigate LLM-assisted cheating in crowdsourcing with a focus on annotation tasks. Our approach quantifies the correlations between worker answers while conditioning on (a subset of) LLM-generated labels available to the requester. Building on prior research, we propose a training-free scoring mechanism with theoretical guarantees under a crowdsourcing model that accounts for LLM collusion. We establish conditions under which our method is effective and empirically demonstrate its robustness in detecting low-effort cheating on real-world crowdsourcing datasets.
Authors: Yutao Jin, Haowen Xiao, Junyong Zhai, Yuxiao Li, Jielei Chu, Fengmao Lv, Yuxiao Li
Abstract: Mild Cognitive Impairment (MCI) serves as a prodromal stage of Alzheimer's Disease (AD), where early identification and intervention can effectively slow the progression to dementia. However, diagnosing AD remains a significant challenge in neurology due to the confounders caused mainly by the selection bias of multi-modal data and the complex relationships between variables. To address these issues, we propose a novel visual-language causality-inspired framework named Cross-modal Causal Intervention with Mediator for Alzheimer's Disease Diagnosis (MediAD) for diagnostic assistance. Our MediAD employs Large Language Models (LLMs) to summarize clinical data under strict templates, therefore enriching textual inputs. The MediAD model utilizes Magnetic Resonance Imaging (MRI), clinical data, and textual data enriched by LLMs to classify participants into Cognitively Normal (CN), MCI, and AD categories. Because of the presence of confounders, such as cerebral vascular lesions and age-related biomarkers, non-causal models are likely to capture spurious input-output correlations, generating less reliable results. Our framework implicitly mitigates the effect of both observable and unobservable confounders through a unified causal intervention method. Experimental results demonstrate the outstanding performance of our method in distinguishing CN/MCI/AD cases, outperforming other methods in most evaluation metrics. The study showcases the potential of integrating causal reasoning with multi-modal learning for neurological disease diagnosis.
Authors: Yoonjin Lee, Munhee Kim, Hanbi Choi, Juhyeon Park, Seungho Lyoo, Woojin Park
Abstract: Despite advances in financial AI, the automation of evidence-based reasoning remains unresolved in corporate credit assessment, where qualitative non-financial indicators exert decisive influence on loan repayment outcomes yet resist formalization. Existing approaches focus predominantly on numerical prediction and provide limited support for the interpretive judgments required in professional loan evaluation. This study develops and evaluates two operational large language model (LLM)-based systems designed to generate structured reasoning from non-financial evidence. The first is a non-adversarial single-agent system (NAS) that produces bidirectional analysis through a single-pass reasoning pipeline. The second is a debate-based multi-agent system (KPD-MADS) that operationalizes adversarial verification through a ten-step structured interaction protocol grounded in Karl Popper's critical dialogue framework. Both systems were applied to three real corporate cases and evaluated by experienced credit risk professionals. Compared to manual expert reporting, both systems achieved substantial productivity gains (NAS: 11.55 s per case; KPD-MADS: 91.97 s; human baseline: 1920 s). The KPD-MADS demonstrated superior reasoning quality, receiving higher median ratings in explanatory adequacy (4.0 vs. 3.0), practical applicability (4.0 vs. 3.0), and usability (62.5 vs. 52.5). These findings show that structured multi-agent interaction can enhance reasoning rigor and interpretability in financial AI, advancing scalable and defensible automation in corporate credit assessment.
Authors: Anjie Liu, Jianhong Wang, Samuel Kaski, Jun Wang, Mengyue Yang
Abstract: Steering cooperative multi-agent reinforcement learning (MARL) towards desired outcomes is challenging, particularly when the global guidance from a human on the whole multi-agent system is impractical in a large-scale MARL. On the other hand, designing external mechanisms (e.g., intrinsic rewards and human feedback) to coordinate agents mostly relies on empirical studies, lacking a easy-to-use research tool. In this work, we employ multi-agent influence diagrams (MAIDs) as a graphical framework to address the above issues. First, we introduce the concept of MARL interaction paradigms (orthogonal to MARL learning paradigms), using MAIDs to analyze and visualize both unguided self-organization and global guidance mechanisms in MARL. Then, we design a new MARL interaction paradigm, referred to as the targeted intervention paradigm that is applied to only a single targeted agent, so the problem of global guidance can be mitigated. In implementation, we introduce a causal inference technique, referred to as Pre-Strategy Intervention (PSI), to realize the targeted intervention paradigm. Since MAIDs can be regarded as a special class of causal diagrams, a composite desired outcome that integrates the primary task goal and an additional desired outcome can be achieved by maximizing the corresponding causal effect through the PSI. Moreover, the bundled relevance graph analysis of MAIDs provides a tool to identify whether an MARL learning paradigm is workable under the design of an MARL interaction paradigm. In experiments, we demonstrate the effectiveness of our proposed targeted intervention, and verify the result of relevance graph analysis.
Authors: Patterson Hsieh, Jerry Yeh, Mao-Chi He, Wen-Han Hsieh, Elvis Hsieh
Abstract: Climate change is intensifying the occurrence of harmful algal bloom (HAB), particularly cyanobacteria, which threaten aquatic ecosystems and human health through oxygen depletion, toxin release, and disruption of marine biodiversity. Traditional monitoring approaches, such as manual water sampling, remain labor-intensive and limited in spatial and temporal coverage. Recent advances in vision-language models (VLMs) for remote sensing have shown potential for scalable AI-driven solutions, yet challenges remain in reasoning over imagery and quantifying bloom severity. In this work, we introduce ALGae Observation and Segmentation (ALGOS), a segmentation-and-reasoning system for HAB monitoring that combines remote sensing image understanding with severity estimation. Our approach integrates GeoSAM-assisted human evaluation for high-quality segmentation mask curation and fine-tunes vision language model on severity prediction using the Cyanobacteria Aggregated Manual Labels (CAML) from NASA. Experiments demonstrate that ALGOS achieves robust performance on both segmentation and severity-level estimation, paving the way toward practical and automated cyanobacterial monitoring systems.
Authors: Hamid R. Tizhoosh
Abstract: In non-medical domains, foundation models (FMs) have revolutionized computer vision and language processing through large-scale self-supervised and multimodal learning. Consequently, their rapid adoption in computational pathology was expected to deliver comparable breakthroughs in cancer diagnosis, prognostication, and multimodal retrieval. However, recent systematic evaluations reveal fundamental weaknesses: low diagnostic accuracy, poor robustness, geometric instability, heavy computational demands, and concerning safety vulnerabilities. This short paper examines these shortcomings and argues that they stem from deeper conceptual mismatches between the assumptions underlying generic foundation modeling in mainstream AI and the intrinsic complexity of human tissue. Seven interrelated causes are identified: biological complexity, ineffective self-supervision, overgeneralization, excessive architectural complexity, lack of domain-specific innovation, insufficient data, and a fundamental design flaw related to tissue patch size. These findings suggest that current pathology foundation models remain conceptually misaligned with the nature of tissue morphology and call for a fundamental rethinking of the paradigm itself.
Authors: Qianli Shen, Daoyuan Chen, Yilun Huang, Zhenqing Ling, Yaliang Li, Bolin Ding, Jingren Zhou
Abstract: Reinforcement finetuning (RFT) is a key technique for aligning Large Language Models (LLMs) with human preferences and enhancing reasoning, yet its effectiveness is highly sensitive to which tasks are explored during training. Uniform task sampling is inefficient, wasting computation on tasks that are either trivial or unsolvable, while existing task selection methods often suffer from high rollout costs, poor adaptivity, or incomplete evidence. We introduce BOTS, a unified framework for Bayesian Online Task Selection in LLM reinforcement finetuning. Grounded in Bayesian inference, BOTS adaptively maintains posterior estimates of task difficulty as the model evolves. It jointly incorporates explicit evidence from direct evaluations of selected tasks and implicit evidence inferred from these evaluations for unselected tasks, with Thompson sampling ensuring a principled balance between exploration and exploitation. To make implicit evidence practical, we instantiate it with an ultra-light interpolation-based plug-in that estimates difficulties of unevaluated tasks without extra rollouts, adding negligible overhead. Empirically, across diverse domains and LLM scales, BOTS consistently improves data efficiency and performance over baselines and ablations, providing a practical and extensible solution for dynamic task selection in RFT.
Authors: Mohamed Bouadi, Pratinav Seth, Aditya Tanna, Vinay Kumar Sankarapu
Abstract: Tabular data remain the predominant format for real-world applications. Yet, developing effective neural models for tabular data remains challenging due to heterogeneous feature types and complex interactions occurring at multiple scales. Recent advances in tabular in-context learning (ICL), such as TabPFN and TabICL, have achieved state-of-the-art performance comparable to gradient-boosted trees (GBTs) without task-specific fine-tuning. However, current architectures exhibit key limitations: (1) single-scale feature processing that overlooks hierarchical dependencies, (2) dense attention with quadratic scaling in table width, and (3) strictly sequential component processing that prevents iterative representation refinement and cross-component communication. To address these challenges, we introduce Orion-MSP, a tabular ICL architecture featuring three key innovations: (1) multi-scale processing to capture hierarchical feature interactions; (2) block-sparse attention combining windowed, global, and random patterns for scalable efficiency and long-range connectivity; and (3) a Perceiver-style memory enabling safe bidirectional information flow across components. Across diverse benchmarks, Orion-MSP matches or surpasses state-of-the-art performance while scaling effectively to high-dimensional tables, establishing a new standard for efficient tabular in-context learning. The model is publicly available at https://github.com/Lexsi-Labs/Orion-MSP .
Authors: Jonathan Li, Nasim Farahini, Evgenii Iuliugin, Magnus Vesterlund, Christian Haggstrom, Guangtao Wang, Shubhangi Upasani, Ayush Sachdeva, Rui Li, Faline Fu, Chen Wu, Ayesha Siddiqua, John Long, Tuowen Zhao, Matheen Musaddiq, Hakan Zeffer, Yun Du, Mingran Wang, Qinghua Li, Bo Li, Urmish Thakker, Raghu Prabhakar
Abstract: The proliferation of 100B+ parameter Large Language Models (LLMs) with 100k+ context length support have resulted in increasing demands for on-chip memory to support large KV caches. Techniques such as StreamingLLM and SnapKV demonstrate how to control KV cache size while maintaining model accuracy. Yet, these techniques are not commonly used within industrial deployments using frameworks like vLLM or SGLang. The reason is twofold: on one hand, the static graphs and continuous batching methodology employed by these frameworks make it difficult to admit modifications to the standard multi-head attention algorithm, while on the other hand, the accuracy implications of such techniques on modern instruction-following and reasoning models are not well understood, obfuscating the need for implementing these techniques. In this paper, we explore these accuracy implications on Llama-3.1-8B-Instruct and DeepSeek-R1, and develop SnapStream, a KV cache compression method that can be deployed at scale. We demonstrate the efficacy of SnapStream in a 16-way tensor-parallel deployment of DeepSeek-671B on SambaNova SN40L accelerators running at 128k context length and up to 1832 tokens per second in a real production setting. SnapStream enables $4\times$ improved on-chip memory usage and introduces minimal accuracy degradation on LongBench-v2, AIME24 and LiveCodeBench. To the best of our knowledge, this is the first implementation of sparse KV attention techniques deployed in a production inference system with static graphs and continuous batching.
Authors: Varun Kumar, George Em Karniadakis
Abstract: The engineering design process often demands expertise from multiple domains, leading to complex collaborations and iterative refinements. Traditional methods can be resource-intensive and prone to inefficiencies. To address this, we formalize the engineering design process through a multi-agent AI framework that integrates structured design and review loops. The framework introduces specialized knowledge-driven agents that collaborate to generate and refine design candidates. As an exemplar, we demonstrate its application to the aerodynamic optimization of 4-digit NACA airfoils. The framework consists of three key AI agents: a Graph Ontologist, a Design Engineer, and a Systems Engineer. The Graph Ontologist employs a Large Language Model (LLM) to construct two domain-specific knowledge graphs from airfoil design literature. The Systems Engineer, informed by a human manager, formulates technical requirements that guide design generation and evaluation. The Design Engineer leverages the design knowledge graph and computational tools to propose candidate airfoils meeting these requirements. The Systems Engineer reviews and provides feedback both qualitative and quantitative using its own knowledge graph, forming an iterative feedback loop until a design is validated by the manager. The final design is then optimized to maximize performance metrics such as the lift-to-drag ratio. Overall, this work demonstrates how collaborative AI agents equipped with structured knowledge representations can enhance efficiency, consistency, and quality in the engineering design process.
Authors: Shao-Qun Zhang, Zong-Yi Chen, Yong-Ming Tian, Xun Lu
Abstract: Past decades have witnessed a great interest in the distinction and connection between neural network learning and kernel learning. Recent advancements have made theoretical progress in connecting infinite-wide neural networks and Gaussian processes. Two predominant approaches have emerged: the Neural Network Gaussian Process (NNGP) and the Neural Tangent Kernel (NTK). The former, rooted in Bayesian inference, represents a zero-order kernel, while the latter, grounded in the tangent space of gradient descents, is a first-order kernel. In this paper, we present the Unified Neural Kernel (UNK), which {is induced by the inner product of produced variables and characterizes the learning dynamics of neural networks with gradient descents and parameter initialization.} The proposed UNK kernel maintains the limiting properties of both NNGP and NTK, exhibiting behaviors akin to NTK with a finite learning step and converging to NNGP as the learning step approaches infinity. Besides, we also theoretically characterize the uniform tightness and learning convergence of the UNK kernel, providing comprehensive insights into this unified kernel. Experimental results underscore the effectiveness of our proposed method.
Authors: Yuansan Liu, Sudanthi Wijewickrema, Dongting Hu, Christofer Bester, Stephen O'Leary, James Bailey
Abstract: Recent innovations in diffusion probabilistic models have paved the way for significant progress in image, text and audio generation, leading to their applications in generative time series forecasting. However, leveraging such abilities to model highly stochastic time series data remains a challenge. In this paper, we propose a novel Stochastic Diffusion (StochDiff) model which learns data-driven prior knowledge at each time step by utilizing the representational power of the stochastic latent spaces to model the variability of the multivariate time series data. The learnt prior knowledge helps the model to capture complex temporal dynamics and the inherent uncertainty of the data. This improves its ability to model highly stochastic time series data. Through extensive experiments on real-world datasets, we demonstrate the effectiveness of our proposed model on stochastic time series forecasting. Additionally, we showcase an application of our model for real-world surgical guidance, highlighting its potential to benefit the medical community.
Authors: Emanuele Zappala
Abstract: We obtain a new universal approximation theorem for continuous (possibly nonlinear) operators on arbitrary Banach spaces using the Leray-Schauder mapping. Moreover, we introduce and study a method for operator learning in Banach spaces $L^p$ of functions with multiple variables, based on orthogonal projections on polynomial bases. We derive a universal approximation result for operators where we learn a linear projection and a finite dimensional mapping under some additional assumptions. For the case of $p=2$, we give some sufficient conditions for the approximation results to hold. This article serves as the theoretical framework for a deep learning methodology in operator learning.
Authors: Elinor Poole-Dayan, Deb Roy, Jad Kabbara
Abstract: While state-of-the-art large language models (LLMs) have shown impressive performance on many tasks, there has been extensive research on undesirable model behavior such as hallucinations and bias. In this work, we investigate how the quality of LLM responses changes in terms of information accuracy, truthfulness, and refusals depending on three user traits: English proficiency, education level, and country of origin. We present extensive experimentation on three state-of-the-art LLMs and two different datasets targeting truthfulness and factuality. Our findings suggest that undesirable behaviors in state-of-the-art LLMs occur disproportionately more for users with lower English proficiency, of lower education status, and originating from outside the US, rendering these models unreliable sources of information towards their most vulnerable users.
Authors: Junkai Liu, Yujie Tong, Hui Huang, Bowen Zheng, Yiran Hu, Peicheng Wu, Chuan Xiao, Makoto Onizuka, Muyun Yang, Shuyuan Zheng
Abstract: Legal judgment prediction (LJP), which enables litigants and their lawyers to forecast judgment outcomes and refine litigation strategies, has emerged as a crucial legal NLP task. Existing studies typically utilize legal facts, i.e., facts that have been established by evidence and determined by the judge, to predict the judgment. However, legal facts are often difficult to obtain in the early stages of litigation, significantly limiting the practical applicability of fact-based LJP. To address this limitation, we propose a novel legal NLP task: legal fact prediction (LFP), which takes the evidence submitted by litigants for trial as input to predict legal facts, thereby empowering fact-based LJP technologies to make predictions in the absence of ground-truth legal facts. We also propose the first benchmark dataset, LFPBench, for evaluating the LFP task. Our extensive experiments on LFPBench demonstrate the effectiveness of LFP-empowered LJP and highlight promising research directions for LFP.
Authors: Ray Congrui Yu, Sherry Wu, Jiang Gui
Abstract: Despite their immense success, deep convolutional neural networks (CNNs) can be difficult to optimize and costly to train due to hundreds of layers within the network depth. Conventional convolutional operations are fundamentally limited by their linear nature along with fixed activations, where many layers are needed to learn meaningful patterns in data. Because of the sheer size of these networks, this approach is simply computationally inefficient, and poses overfitting or gradient explosion risks, especially in small datasets. As a result, we introduce a "plug-in" module, called Residual Kolmogorov-Arnold Network (RKAN). Our module is highly compact, so it can be easily added into any stage (level) of traditional deep networks, where it learns to integrate supportive polynomial feature transformations to existing convolutional frameworks. RKAN offers consistent improvements over baseline models in different vision tasks and widely tested benchmarks, accomplishing cutting-edge performance on them.
Authors: Nithin Somasekharan, Shaowu Pan
Abstract: Representation learning for high-dimensional, complex physical systems aims to identify a low-dimensional intrinsic latent space, which is crucial for reduced-order modeling and modal analysis. To overcome the well-known Kolmogorov barrier, deep autoencoders (AEs) have been introduced in recent years, but they often suffer from poor convergence behavior as the rank of the latent space increases. To address this issue, we propose the learnable weighted hybrid autoencoder, a hybrid approach that combines the strengths of singular value decomposition (SVD) with deep autoencoders through a learnable weighted framework. We find that the introduction of learnable weighting parameters is essential -- without them, the resulting model would either collapse into a standard POD or fail to exhibit the desired convergence behavior. Interestingly, we empirically find that our trained model has a sharpness thousands of times smaller compared to other models. Our experiments on classical chaotic PDE systems, including the 1D Kuramoto-Sivashinsky and forced isotropic turbulence datasets, demonstrate that our approach significantly improves generalization performance compared to several competing methods. Additionally, when combining with time series modeling techniques (e.g., Koopman operator, LSTM), the proposed technique offers significant improvements for surrogate modeling of high-dimensional multi-scale PDE systems.
Authors: Tianyue H. Zhang, Lucas Maes, Alan Milligan, Alexia Jolicoeur-Martineau, Ioannis Mitliagkas, Damien Scieur, Simon Lacoste-Julien, Charles Guille-Escuret
Abstract: Despite its widespread adoption, Adam's advantage over Stochastic Gradient Descent (SGD) lacks a comprehensive theoretical explanation. This paper investigates Adam's sensitivity to rotations of the parameter space. We observe that Adam's performance in training transformers degrades under random rotations of the parameter space, indicating a crucial sensitivity to the choice of basis in practice. This reveals that conventional rotation-invariant assumptions are insufficient to capture Adam's advantages theoretically. To better understand the rotation-dependent properties that benefit Adam, we also identify structured rotations that preserve or even enhance its empirical performance. We then examine the rotation-dependent assumptions in the literature and find that they fall short in explaining Adam's behaviour across various rotation types. In contrast, we verify the orthogonality of the update as a promising indicator of Adam's basis sensitivity, suggesting it may be the key quantity for developing rotation-dependent theoretical frameworks that better explain its empirical success.
Authors: Yiqi Yang, Hongye Fu
Abstract: We present a novel black-box jailbreaking framework that integrates multiple LLM-as-Attacker strategies to deliver highly transferable and effective attacks. The framework is grounded in three key insights from prior jailbreaking research and practice: ensemble approaches outperform single methods in exposing aligned LLM vulnerabilities, malicious instructions vary in jailbreaking difficulty requiring tailored optimization, and disrupting semantic coherence of malicious prompts can manipulate their embeddings to boost success rates. Validated in the Competition for LLM and Agent Safety 2024, our solution achieved top rankings in the Jailbreaking Attack Track.
Authors: Advik Raj Basani, Xiao Zhang
Abstract: LLMs have shown impressive capabilities across various natural language processing tasks, yet remain vulnerable to input prompts, known as jailbreak attacks, carefully designed to bypass safety guardrails and elicit harmful responses. Traditional methods rely on manual heuristics but suffer from limited generalizability. Despite being automatic, optimization-based attacks often produce unnatural prompts that can be easily detected by safety filters or require high computational costs due to discrete token optimization. In this paper, we introduce Generative Adversarial Suffix Prompter (GASP), a novel automated framework that can efficiently generate human-readable jailbreak prompts in a fully black-box setting. In particular, GASP leverages latent Bayesian optimization to craft adversarial suffixes by efficiently exploring continuous latent embedding spaces, gradually optimizing the suffix prompter to improve attack efficacy while balancing prompt coherence via a targeted iterative refinement procedure. Through comprehensive experiments, we show that GASP can produce natural adversarial prompts, significantly improving jailbreak success over baselines, reducing training times, and accelerating inference speed, thus making it an efficient and scalable solution for red-teaming LLMs.
Authors: Matias D. Cattaneo, Boris Shigida
Abstract: In modern optimization methods used in deep learning, each update depends on the history of previous iterations, often referred to as memory, and this dependence decays fast as the iterates go further into the past. For example, gradient descent with momentum has exponentially decaying memory through exponentially averaged past gradients. We introduce a general technique for identifying a memoryless algorithm that approximates an optimization algorithm with memory. It is obtained by replacing all past iterates in the update by the current one, and then adding a correction term arising from memory (also a function of the current iterate). This correction term can be interpreted as a perturbation of the loss, and the nature of this perturbation can inform how memory implicitly (anti-)regularizes the optimization dynamics. As an application of our theory, we find that Lion does not have the kind of implicit anti-regularization induced by memory that AdamW does, providing a theory-based explanation for Lion's better generalization performance recently documented.
Authors: Belinda Mo, Kyssen Yu, Joshua Kazdan, Joan Cabezas, Proud Mpala, Lisa Yu, Chris Cundy, Charilaos Kanatsoulis, Sanmi Koyejo
Abstract: Recent interest in building foundation models for KGs has highlighted a fundamental challenge: knowledge-graph data is relatively scarce. The best-known KGs are primarily human-labeled, created by pattern-matching, or extracted using early NLP techniques. While human-generated KGs are in short supply, automatically extracted KGs are of questionable quality. We present a solution to this data scarcity problem in the form of a text-to-KG generator (KGGen), a package that uses language models to create high-quality graphs from plaintext. Unlike other KG extractors, KGGen clusters related entities to reduce sparsity in extracted KGs. KGGen is available as a Python library (\texttt{pip install kg-gen}), making it accessible to everyone. Along with KGGen, we release the first benchmark, Measure of of Information in Nodes and Edges (MINE), that tests an extractor's ability to produce a useful KG from plain text. We benchmark our new tool against existing extractors and demonstrate far superior performance.
Authors: Zhuchen Cao, Sven Apel, Adish Singla, Vera Demberg
Abstract: Large Language Models (LLMs) have demonstrated impressive potential in translating natural language (NL) instructions into program code. However, user instructions often contain inherent ambiguities, making it challenging for LLMs to generate code that accurately reflects the user's true intent. To address this challenge, researchers have proposed approaches that produce multiple candidates of the program code and then rerank them to identify the best solution. In this paper, we propose CodeRSA, a novel code candidate reranking mechanism built upon the Rational Speech Act (RSA) framework, designed to guide LLMs toward more comprehensive pragmatic reasoning about user intent. We evaluate CodeRSA using Llama-3-8B-Instruct and Qwen-2.5-7B-Instruct on two widely used code generation benchmarks, HumanEval and MBPP. Our experiment results show that CodeRSA consistently outperforms common baselines, surpasses the state-of-the-art approach in most cases, and demonstrates robust overall performance. These findings underscore the effectiveness of integrating pragmatic reasoning into code candidate reranking, offering a promising direction for enhancing code generation quality in LLMs.
Authors: Borong Zhang, Yuhao Zhang, Jiaming Ji, Yingshan Lei, Josef Dai, Yuanpei Chen, Yaodong Yang
Abstract: Vision-language-action models (VLAs) show potential as generalist robot policies. However, these models pose extreme safety challenges during real-world deployment, including the risk of harm to the environment, the robot itself, and humans. How can safety constraints be explicitly integrated into VLAs? We address this by exploring an integrated safety approach (ISA), systematically modeling safety requirements, then actively eliciting diverse unsafe behaviors, effectively constraining VLA policies via safe reinforcement learning, and rigorously assuring their safety through targeted evaluations. Leveraging the constrained Markov decision process (CMDP) paradigm, ISA optimizes VLAs from a min-max perspective against elicited safety risks. Thus, policies aligned through this comprehensive approach achieve the following key features: (I) effective safety-performance trade-offs, reducing the cumulative cost of safety violations by 83.58% compared to the state-of-the-art method, while also maintaining task success rate (+3.85%). (II) strong safety assurance, with the ability to mitigate long-tail risks and handle extreme failure scenarios. (III) robust generalization of learned safety behaviors to various out-of-distribution perturbations. The effectiveness is evaluated on long-horizon mobile manipulation tasks. Our data, models and newly proposed benchmark environment are available at https://pku-safevla.github.io.
Authors: Pin-Jie Lin, Rishab Balasubramanian, Fengyuan Liu, Nikhil Kandpal, Tu Vu
Abstract: Modern LLMs struggle with efficient updates, as each new pretrained model version requires repeating expensive alignment processes. This challenge also applies to domain- or languagespecific models, where fine-tuning on specialized data must be redone for every new base model release. In this paper, we explore the transfer of fine-tuning updates between model versions. Specifically, we derive the diff vector (representing the weight changes from finetuning) from one source model version and apply it to the base model of a different target version. Through empirical evaluations on various open-weight model versions, we show that transferring diff vectors can significantly improve the performance of the target base model. For example, transferring the fine-tuning updates from Llama 3.0 8B improves Llama 3.1 8B by 46.9% on IFEval and 15.7% on LiveCodeBench without additional training, even surpassing Llama 3.1 8B Instruct. Furthermore, we demonstrate performance gains on multilingual tasks, with 4.7% and 15.5% improvements on Global MMLU for Malagasy and Turkish, respectively. We observe that these merged models provide stronger initializations for further fine-tuning. Lastly, our controlled experiments suggest that fine-tuning transfer is most effective when source and target models lie in a linearly connected region of parameter space, and we provide a theoretical analysis of our method. Taken together, fine-tuning transfer offers a cost-efficient and practical strategy for continuous LLM development. Our code is available at github.com/pjlintw/finetuning-transfer.
Authors: Hung-Yueh Chiang, Chi-Chih Chang, Natalia Frumkin, Kai-Chiang Wu, Mohamed S. Abdelfattah, Diana Marculescu
Abstract: State Space Models (SSMs) are emerging as a compelling alternative to Transformers because of their consistent memory usage and high performance. Despite this, scaling up SSMs on cloud services or limited-resource devices is challenging due to their storage requirements and computational power. To overcome this, quantizing SSMs with low bit-width data formats can reduce model size and benefit from hardware acceleration. As SSMs are prone to quantization-induced errors, recent efforts have focused on optimizing a particular model or bit-width for efficiency without sacrificing performance. However, distinct bit-width configurations are essential for different scenarios, like W4A8 for boosting large-batch decoding speed, and W4A16 for enhancing generation speed in short prompt applications for a single user. To this end, we present Quamba2, compatible with W8A8, W4A8, and W4A16 for both Mamba1 and Mamba2 backbones, addressing the growing demand for SSM deployment on various platforms. Based on the channel order preserving and activation persistence of SSMs, we propose an offline approach to quantize inputs of a linear recurrence in 8-bit by sorting and clustering for input $x$, combined with a per-state-group quantization for input-dependent parameters $B$ and $C$. To ensure compute-invariance in the SSM output, we rearrange weights offline according to the clustering sequence. The experiments show that Quamba2-8B outperforms two state-of-the-art SSM quantization methods and delivers 1.3$\times$ and 3$\times$ speed-ups in the pre-filling and generation stages, respectively, while offering 4$\times$ memory reduction with only a $1.6\%$ average accuracy drop. The evaluation on MMLU shows the generalizability and robustness of our framework. The code and quantized models will be released at: https://github.com/enyac-group/Quamba.
Authors: Shubham Kumar Nigam, Balaramamahanthi Deepak Patnaik, Shivam Mishra, Noel Shallum, Kripabandhu Ghosh, Arnab Bhattacharya
Abstract: In the landscape of Fact-based Judgment Prediction and Explanation (FJPE), reliance on factual data is essential for developing robust and realistic AI-driven decision-making tools. This paper introduces TathyaNyaya, the largest annotated dataset for FJPE tailored to the Indian legal context, encompassing judgments from the Supreme Court of India and various High Courts. Derived from the Hindi terms "Tathya" (fact) and "Nyaya" (justice), the TathyaNyaya dataset is uniquely designed to focus on factual statements rather than complete legal texts, reflecting real-world judicial processes where factual data drives outcomes. Complementing this dataset, we present FactLegalLlama, an instruction-tuned variant of the LLaMa-3-8B Large Language Model (LLM), optimized for generating high-quality explanations in FJPE tasks. Finetuned on the factual data in TathyaNyaya, FactLegalLlama integrates predictive accuracy with coherent, contextually relevant explanations, addressing the critical need for transparency and interpretability in AI-assisted legal systems. Our methodology combines transformers for binary judgment prediction with FactLegalLlama for explanation generation, creating a robust framework for advancing FJPE in the Indian legal domain. TathyaNyaya not only surpasses existing datasets in scale and diversity but also establishes a benchmark for building explainable AI systems in legal analysis. The findings underscore the importance of factual precision and domain-specific tuning in enhancing predictive performance and interpretability, positioning TathyaNyaya and FactLegalLlama as foundational resources for AI-assisted legal decision-making.
Authors: Manveer Singh Tamber, Forrest Sheng Bao, Chenyu Xu, Ge Luo, Suleman Kazi, Minseok Bae, Miaoran Li, Ofer Mendelevitch, Renyi Qu, Jimmy Lin
Abstract: Retrieval-augmented generation (RAG) aims to reduce hallucinations by grounding responses in external context, yet large language models (LLMs) still frequently introduce unsupported information or contradictions even when provided with relevant context. This paper presents two complementary efforts at Vectara to measure and benchmark LLM faithfulness in RAG. First, we describe our original hallucination leaderboard, which has tracked hallucination rates for LLMs since 2023 using our HHEM hallucination detection model. Motivated by limitations observed in current hallucination detection methods, we introduce FaithJudge, an LLM-as-a-judge framework that leverages a pool of diverse human-annotated hallucination examples to substantially improve the automated hallucination evaluation of LLMs. We introduce an enhanced hallucination leaderboard centered on FaithJudge that benchmarks LLMs on RAG faithfulness in summarization, question-answering, and data-to-text generation tasks. FaithJudge enables a more reliable benchmarking of LLM hallucinations in RAG and supports the development of more trustworthy generative AI systems: https://github.com/vectara/FaithJudge.
Authors: Steven Song, Morgan Borjigin-Wang, Irene Madejski, Robert L. Grossman
Abstract: The Cancer Genome Atlas (TCGA) has enabled novel discoveries and served as a large-scale reference dataset in cancer through its harmonized genomics, clinical, and imaging data. Numerous prior studies have developed bespoke deep learning models over TCGA for tasks such as cancer survival prediction. A modern paradigm in biomedical deep learning is the development of foundation models (FMs) to derive feature embeddings agnostic to a specific modeling task. Biomedical text especially has seen growing development of FMs. While TCGA contains free-text data as pathology reports, these have been historically underutilized. Here, we investigate the ability to train classical machine learning models over multimodal, zero-shot FM embeddings of cancer data. We demonstrate the ease and additive effect of multimodal fusion, outperforming unimodal models. Further, we show the benefit of including pathology report text and rigorously evaluate the effect of model-based text summarization and hallucination. Overall, we propose an embedding-centric approach to multimodal cancer modeling.
Authors: Giyeong Oh, Woohyun Cho, Siyeol Kim, Suhwan Choi, Youngjae Yu
Abstract: Residual connections are pivotal for deep neural networks, enabling greater depth by mitigating vanishing gradients. However, in standard residual updates, the module's output is directly added to the input stream. This can lead to updates that predominantly reinforce or modulate the existing stream direction, potentially underutilizing the module's capacity for learning entirely novel features. In this work, we introduce Orthogonal Residual Update: we decompose the module's output relative to the input stream and add only the component orthogonal to this stream. This design aims to guide modules to contribute primarily new representational directions, fostering richer feature learning while promoting more efficient training. We demonstrate that our orthogonal update strategy improves generalization accuracy and training stability across diverse architectures (ResNetV2, Vision Transformers) and datasets (CIFARs, TinyImageNet, ImageNet-1k), achieving, for instance, a +3.78 pp top-1 accuracy gain for ViT-B on ImageNet-1k.
Authors: Yoav Ger, Omri Barak
Abstract: Recurrent neural networks (RNNs) trained on neuroscience-inspired tasks offer powerful models of brain computation. However, typical training paradigms rely on open-loop, supervised settings, whereas real-world learning unfolds in closed-loop environments. Here, we develop a mathematical theory describing the learning dynamics of linear RNNs trained in closed-loop contexts. We first demonstrate that two otherwise identical RNNs, trained in either closed- or open-loop modes, follow markedly different learning trajectories. To probe this divergence, we analytically characterize the closed-loop case, revealing distinct stages aligned with the evolution of the training loss. Specifically, we show that the learning dynamics of closed-loop RNNs, in contrast to open-loop ones, are governed by an interplay between two competing objectives: short-term policy improvement and long-term stability of the agent-environment interaction. Finally, we apply our framework to a realistic motor control task, highlighting its broader applicability. Taken together, our results underscore the importance of modeling closed-loop dynamics in a biologically plausible setting.
Authors: Leon Eshuijs, Archie Chaudhury, Alan McBeth, Ethan Nguyen
Abstract: Detecting subtle forms of dishonesty like sycophancy and manipulation in Large Language Models (LLMs) remains challenging for both humans and automated evaluators, as these behaviors often appear through small biases rather than clear false statements. We introduce Judge Using Safety-Steered Alternatives (JUSSA), a novel framework that employs steering vectors not to improve model behavior directly, but to enhance LLM judges' evaluation capabilities. JUSSA applies steering vectors during inference to generate more honest alternatives, providing judges with contrastive examples that make subtle dishonest patterns easier to detect. While existing evaluation methods rely on black-box evaluation, JUSSA leverages model internals to create targeted comparisons from single examples. We evaluate our method on sycophancy detection and introduce a new manipulation dataset covering multiple types of manipulation. Our results demonstrate that JUSSA effectively improves detection accuracy over single-response evaluation in various cases. Analysis across judge models reveals that JUSSA helps weaker judges on easier dishonesty detection tasks, and stronger judges on harder tasks. Layer-wise experiments show how dishonest prompts cause representations to diverge from honest ones in middle layers, revealing where steering interventions are most effective for generating contrastive examples. By demonstrating that steering vectors can enhance safety evaluation rather than just modify behavior, our work opens new directions for scalable model auditing as systems become increasingly sophisticated.
Authors: Charles Hong, Sahil Bhatia, Alvin Cheung, Yakun Sophia Shao
Abstract: Hardware accelerators, especially those designed for tensor processing, have become ubiquitous in today's computing landscape. However, even with significant efforts in building compilers, programming these tensor accelerators remains challenging, leaving much of their potential underutilized. Recently, large language models (LLMs), trained on large amounts of code, have shown significant promise in code generation and optimization tasks, but generating low-resource languages, such as specialized tensor accelerator code still poses a significant challenge. We tackle this challenge with Autocomp, an approach that empowers accelerator programmers to leverage domain knowledge and hardware feedback to optimize code via an automated LLM-driven search. We accomplish this by: 1) formulating each optimization pass as a structured two-phase prompt, divided into planning and code generation phases, 2) inserting domain knowledge during planning via a concise and adaptable optimization menu, and 3) integrating correctness and performance metrics from hardware as feedback at each search iteration. Across three distinct hardware platforms, we demonstrate that Autocomp-optimized code runs 5.6x faster than the vendor-provided library (Gemmini), outperforms expert-level hand-tuned code by 1.9x (AWS Trainium), and achieves 3.8x higher performance than a machine learning-based cost model for GPUs (NVIDIA L40S). Additionally, we demonstrate that optimization schedules generated from Autocomp can be reused across similar tensor operations, improving speedups by up to 24% under a fixed sample budget.
Authors: Pankaj Kumar, Subhankar Mishra
Abstract: Large Language Models (LLMs) have emerged as a promising cornerstone for the development of natural language processing (NLP) and artificial intelligence (AI). However, ensuring the robustness of LLMs remains a critical challenge. To address these challenges and advance the field, this survey provides a comprehensive overview of current studies in this area. First, we systematically examine the nature of robustness in LLMs, including its conceptual foundations, the importance of consistent performance across diverse inputs, and the implications of failure modes in real-world applications. Next, we analyze the sources of non-robustness, categorizing intrinsic model limitations, data-driven vulnerabilities, and external adversarial factors that compromise reliability. Following this, we review state-of-the-art mitigation strategies, and then we discuss widely adopted benchmarks, emerging metrics, and persistent gaps in assessing real-world reliability. Finally, we synthesize findings from existing surveys and interdisciplinary studies to highlight trends, unresolved issues, and pathways for future research.
Authors: Miaoyu Li, Qin Chao, Boyang Li
Abstract: Properly evaluating the ability of Video-Language Models (VLMs) to understand long videos remains a challenge. We propose a long-context video understanding benchmark, Causal2Needles, that assesses two crucial abilities insufficiently addressed by existing benchmarks: (1) extracting information from two separate locations (two needles) in a long video and understanding them jointly, and (2) modeling the world in terms of cause and effect in human behaviors. Causal2Needles evaluates these abilities using noncausal one-needle, causal one-needle, and causal two-needle questions. The most complex question type, causal two-needle questions, require extracting information from both the cause and effect events from a long video and the associated narration text. To prevent textual bias, we introduce two complementary question formats: locating the video clip containing the answer, and verbal description of a visual detail from that video clip. Our experiments reveal that models excelling on existing benchmarks struggle with causal 2-needle questions, and the model performance is negatively correlated with the distance between the two needles. These findings highlight critical limitations in current VLMs. The dataset is available at: https://huggingface.co/datasets/causal2needles/Causal2Needles
URLs: https://huggingface.co/datasets/causal2needles/Causal2Needles
Authors: Jaeyoung Choe, Jihoon Kim, Woohwan Jung
Abstract: Retrieval-augmented generation (RAG) based large language models (LLMs) are widely used in finance for their excellent performance on knowledge-intensive tasks. However, standardized documents (e.g., SEC filing) share similar formats such as repetitive boilerplate texts, and similar table structures. This similarity forces traditional RAG methods to misidentify near-duplicate text, leading to duplicate retrieval that undermines accuracy and completeness. To address these issues, we propose the Hierarchical Retrieval with Evidence Curation (HiREC) framework. Our approach first performs hierarchical retrieval to reduce confusion among similar texts. It first retrieve related documents and then selects the most relevant passages from the documents. The evidence curation process removes irrelevant passages. When necessary, it automatically generates complementary queries to collect missing information. To evaluate our approach, we construct and release a Large-scale Open-domain Financial (LOFin) question answering benchmark that includes 145,897 SEC documents and 1,595 question-answer pairs. Our code and data are available at https://github.com/deep-over/LOFin-bench-HiREC.
Authors: Lingkai Kong, Haichuan Wang, Tonghan Wang, Guojun Xiong, Milind Tambe
Abstract: Incorporating pre-collected offline data from a source environment can significantly improve the sample efficiency of reinforcement learning (RL), but this benefit is often challenged by discrepancies between the transition dynamics of the source and target environments. Existing methods typically address this issue by penalizing or filtering out source transitions in high dynamics-gap regions. However, their estimation of the dynamics gap often relies on KL divergence or mutual information, which can be ill-defined when the source and target dynamics have disjoint support. To overcome these limitations, we propose CompFlow, a method grounded in the theoretical connection between flow matching and optimal transport. Specifically, we model the target dynamics as a conditional flow built upon the output distribution of the source-domain flow, rather than learning it directly from a Gaussian prior. This composite structure offers two key advantages: (1) improved generalization for learning target dynamics, and (2) a principled estimation of the dynamics gap via the Wasserstein distance between source and target transitions. Leveraging our principled estimation of the dynamics gap, we further introduce an optimistic active data collection strategy that prioritizes exploration in regions of high dynamics gap, and theoretically prove that it reduces the performance disparity with the optimal policy. Empirically, CompFlow outperforms strong baselines across several RL benchmarks with shifted dynamics.
Authors: Junyi Li, Hwee Tou Ng
Abstract: Large language models (LLMs) have significantly advanced in reasoning tasks through reinforcement learning (RL) optimization, achieving impressive capabilities across various challenging benchmarks. However, our empirical analysis reveals a critical drawback: reasoning-oriented RL fine-tuning significantly increases the prevalence of hallucinations. We theoretically analyze the RL training dynamics, identifying high-variance gradient, entropy-induced randomness, and susceptibility to spurious local optima as key factors leading to hallucinations. To address this drawback, we propose Factuality-aware Step-wise Policy Optimization (FSPO), an innovative RL fine-tuning algorithm incorporating explicit factuality verification at each reasoning step. FSPO leverages automated verification against given evidence to dynamically adjust token-level advantage values, incentivizing factual correctness throughout the reasoning process. Experiments across mathematical reasoning and hallucination benchmarks using Qwen2.5 and Llama models demonstrate that FSPO effectively reduces hallucinations while enhancing reasoning accuracy, substantially improving both reliability and performance.
Authors: Neil He, Rishabh Anand, Hiren Madhu, Ali Maatouk, Smita Krishnaswamy, Leandros Tassiulas, Menglin Yang, Rex Ying
Abstract: Large language models (LLMs) have shown great success in text modeling tasks across domains. However, natural language exhibits inherent semantic hierarchies and nuanced geometric structure, which current LLMs do not capture completely owing to their reliance on Euclidean operations. Recent studies have also shown that not respecting the geometry of token embeddings leads to training instabilities and degradation of generative capabilities. These findings suggest that shifting to non-Euclidean geometries can better align language models with the underlying geometry of text. We thus propose to operate fully in Hyperbolic space, known for its expansive, scale-free, and low-distortion properties. We thus introduce HELM, a family of HypErbolic Large Language Models, offering a geometric rethinking of the Transformer-based LLM that addresses the representational inflexibility, missing set of necessary operations, and poor scalability of existing hyperbolic LMs. We additionally introduce a Mixture-of-Curvature Experts model, HELM-MICE, where each expert operates in a distinct curvature space to encode more fine-grained geometric structure from text, as well as a dense model, HELM-D. For HELM-MICE, we further develop hyperbolic Multi-Head Latent Attention (HMLA) for efficient, reduced-KV-cache training and inference. For both models, we develop essential hyperbolic equivalents of rotary positional encodings and RMS normalization. We are the first to train fully hyperbolic LLMs at billion-parameter scale, and evaluate them on well-known benchmarks such as MMLU and ARC, spanning STEM problem-solving, general knowledge, and commonsense reasoning. Our results show consistent gains from our HELM architectures -- up to 4% -- over popular Euclidean architectures used in LLaMA and DeepSeek, highlighting the efficacy and enhanced reasoning afforded by hyperbolic geometry in large-scale LM pretraining.
Authors: Jos\'e Manuel de Frutos, Manuel A. V\'azquez, Pablo M. Olmos, Joaqu\'in M\'iguez
Abstract: Rank-based statistical metrics, such as the invariant statistical loss (ISL), have recently emerged as robust and practically effective tools for training implicit generative models. In this work, we introduce dual-ISL, a novel likelihood-free objective for training implicit generative models that interchanges the roles of the target and model distributions in the ISL framework, yielding a convex optimization problem in the space of model densities. We prove that the resulting rank-based discrepancy $d_K$ is i) continuous under weak convergence and with respect to the $L^1$ norm, and ii) convex in its first argument-properties not shared by classical divergences such as KL or Wasserstein distances. Building on this, we develop a theoretical framework that interprets $d_K$ as an $L^2$-projection of the density ratio $q = p/\tilde p$ onto a Bernstein polynomial basis, from which we derive exact bounds on the truncation error, precise convergence rates, and a closed-form expression for the truncated density approximation. We further extend our analysis to the multivariate setting via random one-dimensional projections, defining a sliced dual-ISL divergence that retains both convexity and continuity. We empirically show that these theoretical advantages translate into practical ones. Specifically, across several benchmarks dual-ISL converges more rapidly, delivers markedly smoother and more stable training, and more effectively prevents mode collapse than classical ISL and other leading implicit generative methods-while also providing an explicit density approximation.
Authors: Youngwan Lee, Kangsan Kim, Kwanyong Park, Ilcahe Jung, Soojin Jang, Seanie Lee, Yong-Ju Lee, Sung Ju Hwang
Abstract: Despite emerging efforts to enhance the safety of Vision-Language Models (VLMs), current approaches face two main shortcomings. 1) Existing safety-tuning datasets and benchmarks only partially consider how image-text interactions can yield harmful content, often overlooking contextually unsafe outcomes from seemingly benign pairs. This narrow coverage leaves VLMs vulnerable to jailbreak attacks in unseen configurations. 2) Prior methods rely primarily on data-centric tuning, with limited architectural innovations to intrinsically strengthen safety. We address these gaps by introducing a holistic safety dataset and benchmark, \textbf{HoliSafe}, that spans all five safe/unsafe image-text combinations, providing a more robust basis for both training and evaluation (HoliSafe-Bench). We further propose a novel modular framework for enhancing VLM safety with a visual guard module (VGM) designed to assess the harmfulness of input images for VLMs. This module endows VLMs with a dual functionality: they not only learn to generate safer responses but can also provide an interpretable harmfulness classification to justify their refusal decisions. A significant advantage of this approach is its modularity; the VGM is designed as a plug-in component, allowing for seamless integration with diverse pre-trained VLMs across various scales. Experiments show that Safe-VLM with VGM, trained on our HoliSafe, achieves state-of-the-art safety performance across multiple VLM benchmarks. Additionally, the HoliSafe-Bench itself reveals critical vulnerabilities in existing VLM models. We hope that HoliSafe and VGM will spur further research into robust and interpretable VLM safety, expanding future avenues for multimodal alignment.
Authors: Cong Wang, Zexuan Deng, Zhiwei Jiang, Yafeng Yin, Fei Shen, Zifeng Cheng, Shiping Ge, Shiwei Gan, Qing Gu
Abstract: Sign Language Video Generation (SLVG) seeks to generate identity-preserving sign language videos from spoken language texts. Existing methods primarily rely on the single coarse condition (\eg, skeleton sequences) as the intermediary to bridge the translation model and the video generation model, which limits both the naturalness and expressiveness of the generated videos. To overcome these limitations, we propose SignViP, a novel SLVG framework that incorporates multiple fine-grained conditions for improved generation fidelity. Rather than directly translating error-prone high-dimensional conditions, SignViP adopts a discrete tokenization paradigm to integrate and represent fine-grained conditions (\ie, fine-grained poses and 3D hands). SignViP contains three core components. (1) Sign Video Diffusion Model is jointly trained with a multi-condition encoder to learn continuous embeddings that encapsulate fine-grained motion and appearance. (2) Finite Scalar Quantization (FSQ) Autoencoder is further trained to compress and quantize these embeddings into discrete tokens for compact representation of the conditions. (3) Multi-Condition Token Translator is trained to translate spoken language text to discrete multi-condition tokens. During inference, Multi-Condition Token Translator first translates the spoken language text into discrete multi-condition tokens. These tokens are then decoded to continuous embeddings by FSQ Autoencoder, which are subsequently injected into Sign Video Diffusion Model to guide video generation. Experimental results show that SignViP achieves state-of-the-art performance across metrics, including video quality, temporal coherence, and semantic fidelity. The code is available at https://github.com/umnooob/signvip/.
Authors: Xingting Yao, Qinghao Hu, Fei Zhou, Tielong Liu, Gang Li, Peisong Wang, Jian Cheng
Abstract: Multi-bit spiking neural networks (SNNs) have recently become a heated research spot, pursuing energy-efficient and high-accurate AI. However, with more bits involved, the associated memory and computation demands escalate to the point where the performance improvements become disproportionate. Based on the insight that different layers demonstrate different importance and extra bits could be wasted and interfering, this paper presents an adaptive bit allocation strategy for direct-trained SNNs, achieving fine-grained layer-wise allocation of memory and computation resources. Thus, SNN's efficiency and accuracy can be improved. Specifically, we parametrize the temporal lengths and the bit widths of weights and spikes, and make them learnable and controllable through gradients. To address the challenges caused by changeable bit widths and temporal lengths, we propose the refined spiking neuron, which can handle different temporal lengths, enable the derivation of gradients for temporal lengths, and suit spike quantization better. In addition, we theoretically formulate the step-size mismatch problem of learnable bit widths, which may incur severe quantization errors to SNN, and accordingly propose the step-size renewal mechanism to alleviate this issue. Experiments on various datasets, including the static CIFAR and ImageNet datasets and the dynamic CIFAR-DVS, DVS-GESTURE, and SHD datasets, demonstrate that our methods can reduce the overall memory and computation cost while achieving higher accuracy. Particularly, our SEWResNet-34 can achieve a 2.69% accuracy gain and 4.16x lower bit budgets over the advanced baseline work on ImageNet. This work will be open-sourced.
Authors: Xuanming Zhang
Abstract: Industrial carbon emissions are a major driver of climate change, yet modeling these emissions is challenging due to multicollinearity among factors and complex interdependencies across sectors and time. We propose a novel graph-based deep learning framework DGL to analyze and forecast industrial CO_2 emissions, addressing high feature correlation and capturing industrial-temporal interdependencies. Unlike traditional regression or clustering methods, our approach leverages a Graph Neural Network (GNN) with attention mechanisms to model relationships between industries (or regions) and a temporal transformer to learn long-range patterns. We evaluate our framework on public global industry emissions dataset derived from EDGAR v8.0, spanning multiple countries and sectors. The proposed model achieves superior predictive performance - reducing error by over 15% compared to baseline deep models - while maintaining interpretability via attention weights and causal analysis. We believe that we are the first Graph-Temporal architecture that resolves multicollinearity by structurally encoding feature relationships, along with integration of causal inference to identify true drivers of emissions, improving transparency and fairness. We also stand a demonstration of policy relevance, showing how model insights can guide sector-specific decarbonization strategies aligned with sustainable development goals. Based on the above, we show high-emission "hotspots" and suggest equitable intervention plans, illustrating the potential of state-of-the-art AI graph learning to advance climate action, offering a powerful tool for policymakers and industry stakeholders to achieve carbon reduction targets.
Authors: Paulo Mendes, Eva Maia, Isabel Pra\c{c}a
Abstract: Phishing emails continue to pose a significant threat to cybersecurity by exploiting human vulnerabilities through deceptive content and malicious payloads. While Machine Learning (ML) models are effective at detecting phishing threats, their performance largely relies on the quality and diversity of the training data. This paper presents MeAJOR (Merged email Assets from Joint Open-source Repositories) Corpus, a novel, multi-source phishing email dataset designed to overcome critical limitations in existing resources. It integrates 135894 samples representing a broad number of phishing tactics and legitimate emails, with a wide spectrum of engineered features. We evaluated the dataset's utility for phishing detection research through systematic experiments with four classification models (RF, XGB, MLP, and CNN) across multiple feature configurations. Results highlight the dataset's effectiveness, achieving 98.34% F1 with XGB. By integrating broad features from multiple categories, our dataset provides a reusable and consistent resource, while addressing common challenges like class imbalance, generalisability and reproducibility.
Authors: Maxence Bouvier, Ryan Amaudruz, Felix Arnold, Renzo Andri, Lukas Cavigelli
Abstract: As AI workloads proliferate, optimizing arithmetic units is becoming increasingly important for reducing the footprint of digital systems. Conventional design flows, which often rely on manual or heuristic-based optimization, are limited in their ability to thoroughly explore the vast design space. In this paper, we introduce GENIAL, a machine learning-based framework for the automatic generation and optimization of arithmetic units, with a focus on multipliers. At the core of GENIAL is a Transformer-based surrogate model trained in two stages, involving self-supervised pretraining followed by supervised finetuning, to robustly forecast key hardware metrics such as power and area from abstracted design representations. By inverting the surrogate model, GENIAL efficiently searches for new operand encodings that directly minimize power consumption in arithmetic units for specific input data distributions. Extensive experiments on large datasets demonstrate that GENIAL is consistently more sample efficient than other methods, and converges faster towards optimized designs. This enables deployment of a high-effort logic synthesis optimization flow in the loop, improving the accuracy of the surrogate model. Notably, GENIAL automatically discovers encodings that achieve up to 18% switching activity savings within multipliers on representative AI workloads compared with the conventional two's complement. We also demonstrate the versatility of our approach by achieving significant improvements on Finite State Machines, highlighting GENIAL's applicability for a wide spectrum of logic functions. Together, these advances mark a significant step toward automated Quality-of-Results-optimized combinational circuit generation for digital systems.
Authors: Christian Meske, Tobias Hermanns, Esther von der Weiden, Kai-Uwe Loser, Thorsten Berger
Abstract: Software development is undergoing a fundamental transformation as vibe coding becomes widespread, with large portions of contemporary codebases now being AI-generated. The disconnect between rapid adoption and limited conceptual understanding highlights the need for an inquiry into this emerging paradigm. Drawing on an intent perspective and historical analysis, we define vibe coding as a software development paradigm where humans and generative AI engage in collaborative flow to co-create software artifacts through natural language dialogue, shifting the mediation of developer intent from deterministic instruction to probabilistic inference. By intent mediation, we refer to the fundamental process through which developers translate their conceptual goals into representations that computational systems can execute. Our results show that vibe coding reconfigures cognitive work by redistributing epistemic labor between humans and machines, shifting the expertise in the software development process away from traditional areas such as design or technical implementation toward collaborative orchestration. We identify key opportunities, including democratization, acceleration, and systemic leverage, alongside risks, such as black box codebases, responsibility gaps, and ecosystem bias. We conclude with a research agenda spanning human-, technology-, and organization-centered directions to guide future investigations of this paradigm.
Authors: Zhigen Zhao, Liuchuan Yu, Ke Jing, Ning Yang
Abstract: The rapid advancement of Vision-Language-Action models has created an urgent need for large-scale, high-quality robot demonstration datasets. Although teleoperation is the predominant method for data collection, current approaches suffer from limited scalability, complex setup procedures, and suboptimal data quality. This paper presents XRoboToolkit, a cross-platform framework for extended reality based robot teleoperation built on the OpenXR standard. The system features low-latency stereoscopic visual feedback, optimization-based inverse kinematics, and support for diverse tracking modalities including head, controller, hand, and auxiliary motion trackers. XRoboToolkit's modular architecture enables seamless integration across robotic platforms and simulation environments, spanning precision manipulators, mobile robots, and dexterous hands. We demonstrate the framework's effectiveness through precision manipulation tasks and validate data quality by training VLA models that exhibit robust autonomous performance.
Authors: Shubham Kumar Nigam, Balaramamahanthi Deepak Patnaik, Shivam Mishra, Ajay Varghese Thomas, Noel Shallum, Kripabandhu Ghosh, Arnab Bhattacharya
Abstract: Legal Judgment Prediction (LJP) has emerged as a key area in AI for law, aiming to automate judicial outcome forecasting and enhance interpretability in legal reasoning. While previous approaches in the Indian context have relied on internal case content such as facts, issues, and reasoning, they often overlook a core element of common law systems, which is reliance on statutory provisions and judicial precedents. In this work, we propose NyayaRAG, a Retrieval-Augmented Generation (RAG) framework that simulates realistic courtroom scenarios by providing models with factual case descriptions, relevant legal statutes, and semantically retrieved prior cases. NyayaRAG evaluates the effectiveness of these combined inputs in predicting court decisions and generating legal explanations using a domain-specific pipeline tailored to the Indian legal system. We assess performance across various input configurations using both standard lexical and semantic metrics as well as LLM-based evaluators such as G-Eval. Our results show that augmenting factual inputs with structured legal knowledge significantly improves both predictive accuracy and explanation quality.
Authors: Ziye Deng, Ruihan He, Jiaxiang Liu, Yuan Wang, Zijie Meng, Songtao Jiang, Yong Xie, Zuozhu Liu
Abstract: Medical image grounding aims to align natural language phrases with specific regions in medical images, serving as a foundational task for intelligent diagnosis, visual question answering (VQA), and automated report generation (MRG). However, existing research is constrained by limited modality coverage, coarse-grained annotations, and the absence of a unified, generalizable grounding framework. To address these challenges, we construct a large-scale medical grounding dataset Med-GLIP-5M comprising over 5.3 million region-level annotations across seven imaging modalities, covering diverse anatomical structures and pathological findings. The dataset supports both segmentation and grounding tasks with hierarchical region labels, ranging from organ-level boundaries to fine-grained lesions. Based on this foundation, we propose Med-GLIP, a modality-aware grounding framework trained on Med-GLIP-5M. Rather than relying on explicitly designed expert modules, Med-GLIP implicitly acquires hierarchical semantic understanding from diverse training data -- enabling it to recognize multi-granularity structures, such as distinguishing lungs from pneumonia lesions. Extensive experiments demonstrate that Med-GLIP consistently outperforms state-of-the-art baselines across multiple grounding benchmarks. Furthermore, integrating its spatial outputs into downstream tasks, including medical VQA and report generation, leads to substantial performance gains. Our dataset will be released soon.
Authors: Trusting Inekwe, Winnie Mkandawire, Emmanuel Agu, Andres Colubri
Abstract: The COVID-19 pandemic disrupted healthcare systems worldwide, disproportionately impacting individuals with chronic conditions such as cardiovascular disease (CVD). These disruptions -- through delayed care and behavioral changes, affected key CVD biomarkers, including LDL cholesterol (LDL-C), HbA1c, BMI, and systolic blood pressure (SysBP). Accurate modeling of these changes is crucial for predicting disease progression and guiding preventive care. However, prior work has not addressed multi-target prediction of CVD biomarker from Electronic Health Records (EHRs) using machine learning (ML), while jointly capturing biomarker interdependencies, temporal patterns, and predictive uncertainty. In this paper, we propose MBT-CB, a Multi-target Bayesian Transformer (MBT) with pre-trained BERT-based transformer framework to jointly predict LDL-C, HbA1c, BMI and SysBP CVD biomarkers from EHR data. The model leverages Bayesian Variational Inference to estimate uncertainties, embeddings to capture temporal relationships and a DeepMTR model to capture biomarker inter-relationships. We evaluate MBT-CT on retrospective EHR data from 3,390 CVD patient records (304 unique patients) in Central Massachusetts during the Covid-19 pandemic. MBT-CB outperformed a comprehensive set of baselines including other BERT-based ML models, achieving an MAE of 0.00887, RMSE of 0.0135 and MSE of 0.00027, while effectively capturing data and model uncertainty, patient biomarker inter-relationships, and temporal dynamics via its attention and embedding mechanisms. MBT-CB's superior performance highlights its potential to improve CVD biomarker prediction and support clinical decision-making during pandemics.
Authors: Eve Fleisig, Matthias Orlikowski, Philipp Cimiano, Dan Klein
Abstract: For machine learning datasets to accurately represent diverse opinions in a population, they must preserve variation in data labels while filtering out spam or low-quality responses. How can we balance annotator reliability and representation? We empirically evaluate how a range of heuristics for annotator filtering affect the preservation of variation on subjective tasks. We find that these methods, designed for contexts in which variation from a single ground-truth label is considered noise, often remove annotators who disagree instead of spam annotators, introducing suboptimal tradeoffs between accuracy and label diversity. We find that conservative settings for annotator removal (<5%) are best, after which all tested methods increase the mean absolute error from the true average label. We analyze performance on synthetic spam to observe that these methods often assume spam annotators are more random than real spammers tend to be: most spammers are distributionally indistinguishable from real annotators, and the minority that are distinguishable tend to give relatively fixed answers, not random ones. Thus, tasks requiring the preservation of variation reverse the intuition of existing spam filtering methods: spammers tend to be less random than non-spammers, so metrics that assume variation is spam fare worse. These results highlight the need for spam removal methods that account for label diversity.
Authors: Anran Li, Lingfei Qian, Mengmeng Du, Yu Yin, Yan Hu, Zihao Sun, Yihang Fu, Erica Stutz, Xuguang Ai, Qianqian Xie, Rui Zhu, Jimin Huang, Yifan Yang, Siru Liu, Yih-Chung Tham, Lucila Ohno-Machado, Hyunghoon Cho, Zhiyong Lu, Hua Xu, Qingyu Chen
Abstract: Large Language Models (LLMs) have demonstrated significant potential in medicine. To date, LLMs have been widely applied to tasks such as diagnostic assistance, medical question answering, and clinical information synthesis. However, a key open question remains: to what extent do LLMs memorize medical training data. In this study, we present the first comprehensive evaluation of memorization of LLMs in medicine, assessing its prevalence (how frequently it occurs), characteristics (what is memorized), volume (how much content is memorized), and potential downstream impacts (how memorization may affect medical applications). We systematically analyze common adaptation scenarios: (1) continued pretraining on medical corpora, (2) fine-tuning on standard medical benchmarks, and (3) fine-tuning on real-world clinical data, including over 13,000 unique inpatient records from Yale New Haven Health System. The results demonstrate that memorization is prevalent across all adaptation scenarios and significantly higher than reported in the general domain. Memorization affects both the development and adoption of LLMs in medicine and can be categorized into three types: beneficial (e.g., accurate recall of clinical guidelines and biomedical references), uninformative (e.g., repeated disclaimers or templated medical document language), and harmful (e.g., regeneration of dataset-specific or sensitive clinical content). Based on these findings, we offer practical recommendations to facilitate beneficial memorization that enhances domain-specific reasoning and factual accuracy, minimize uninformative memorization to promote deeper learning beyond surface-level patterns, and mitigate harmful memorization to prevent the leakage of sensitive or identifiable patient information.
Authors: Jeffrey Liu, Rongbin Hu
Abstract: Referring Expression Comprehension (REC) is usually addressed with task-trained grounding models. We show that a zero-shot workflow, without any REC-specific training, can achieve competitive or superior performance. Our approach reformulates REC as box-wise visual-language verification: given proposals from a COCO-clean generic detector (YOLO-World), a general-purpose VLM independently answers True/False queries for each region. This simple procedure reduces cross-box interference, supports abstention and multiple matches, and requires no fine-tuning. On RefCOCO, RefCOCO+, and RefCOCOg, our method not only surpasses a zero-shot GroundingDINO baseline but also exceeds reported results for GroundingDINO trained on REC and GroundingDINO+CRG. Controlled studies with identical proposals confirm that verification significantly outperforms selection-based prompting, and results hold with open VLMs. Overall, we show that workflow design, rather than task-specific pretraining, drives strong zero-shot REC performance.
Authors: Nikita Rajaneesh, Thomas Zollo, Richard Zemel
Abstract: Multimodal Large Language Models (MLLMs) hold great promise for advanced reasoning at the intersection of text and images, yet they have not fully realized this potential. MLLMs typically integrate an LLM, a vision encoder, and a connector that maps the vision encoder's embeddings into the LLM's text embedding space. Although each component is pretrained on massive datasets with billions of samples, the entire multimodal model is typically trained on only thousands (or a few million) samples, which can result in weak performance on complex reasoning tasks. To address these shortcomings, instead of relying on extensive labeled datasets for fine-tuning, we propose a Test-Time Warmup method that adapts the MLLM per test instance by leveraging data from weakly supervised auxiliary tasks. With our approach, we observe a relative performance improvement of 4.03% on MMMU, 5.28% on VQA-Rad, and 1.63% on GQA on the Llama-Vision-Instruct model. Our method demonstrates that 'warming up' before inference can enhance MLLMs' robustness across diverse reasoning tasks.
Authors: Kangdi Wang, Zhiyue Wu, Dinghao Zhou, Rui Lin, Junyu Dai, Tao Jiang
Abstract: Variational Autoencoders (VAEs) are essential for large-scale audio tasks like diffusion-based generation. However, existing open-source models often neglect auditory perceptual aspects during training, leading to weaknesses in phase accuracy and stereophonic spatial representation. To address these challenges, we propose {\epsilon}ar-VAE, an open-source music signal reconstruction model that rethinks and optimizes the VAE training paradigm. Our contributions are threefold: (i) A K-weighting perceptual filter applied prior to loss calculation to align the objective with auditory perception. (ii) Two novel phase losses: a Correlation Loss for stereo coherence, and a Phase Loss using its derivatives--Instantaneous Frequency and Group Delay--for precision. (iii) A new spectral supervision paradigm where magnitude is supervised by all four Mid/Side/Left/Right components, while phase is supervised only by the LR components. Experiments show {\epsilon}ar-VAE at 44.1kHz substantially outperforms leading open-source models across diverse metrics, showing particular strength in reconstructing high-frequency harmonics and the spatial characteristics.
Authors: Vaibhav Mishra, William Lotter
Abstract: Foundation models are increasingly developed in computational pathology (CPath) given their promise in facilitating many downstream tasks. While recent studies have evaluated task performance across models, less is known about the structure and variability of their learned representations. Here, we systematically analyze the representational spaces of six CPath foundation models using techniques popularized in computational neuroscience. The models analyzed span vision-language contrastive learning (CONCH, PLIP, KEEP) and self-distillation (UNI (v2), Virchow (v2), Prov-GigaPath) approaches. Through representational similarity analysis using H&E image patches from TCGA, we find that UNI2 and Virchow2 have the most distinct representational structures, whereas Prov-Gigapath has the highest average similarity across models. Having the same training paradigm (vision-only vs. vision-language) did not guarantee higher representational similarity. The representations of all models showed a high slide-dependence, but relatively low disease-dependence. Stain normalization decreased slide-dependence for all models by a range of 5.5% (CONCH) to 20.5% (PLIP). In terms of intrinsic dimensionality, vision-language models demonstrated relatively compact representations, compared to the more distributed representations of vision-only models. These findings highlight opportunities to improve robustness to slide-specific features, inform model ensembling strategies, and provide insights into how training paradigms shape model representations. Our framework is extendable across medical imaging domains, where probing the internal representations of foundation models can support their effective development and deployment.
Authors: Abel Gurung, Joseph Campbell
Abstract: Foundation models excel across diverse tasks, but adapting them to specialized applications often requires fine-tuning, an approach that is memory and compute-intensive. Parameter-efficient fine-tuning (PEFT) methods mitigate this by updating only a small subset of weights. In this paper, we introduce HyperAdapt, a parameter-efficient fine-tuning method that significantly reduces the number of trainable parameters compared to state-of-the-art methods like LoRA. Specifically, HyperAdapt adapts a pre-trained weight matrix by applying row- and column-wise scaling through diagonal matrices, thereby inducing a high-rank update while requiring only $n+m$ trainable parameters for an $n \times m$ matrix. Theoretically, we establish an upper bound on the rank of HyperAdapt's updates, and empirically, we confirm that it consistently induces high-rank transformations across model layers. Experiments on GLUE, arithmetic reasoning, and commonsense reasoning benchmarks with models up to 14B parameters demonstrate that HyperAdapt matches or nearly matches the performance of full fine-tuning and state-of-the-art PEFT methods while using orders of magnitude fewer trainable parameters.
Authors: Jincheng Liu, Sijun He, Jingjing Wu, Xiangsen Wang, Yang Chen, Zhaoqi Kuang, Siqi Bao, Yuan Yao
Abstract: Recent large language models (LLMs) have shown strong reasoning capabilities. However, a critical question remains: do these models possess genuine reasoning skills particularly complex strategic reasoning or are they primarily excelling at sophisticated pattern recognition within their training data? To address this question, this paper presents a chess testbed, ChessArena, to evaluate the strategic reasoning capabilities of LLMs. Chess requires complex strategic reasoning capabilities including long-term planning, strict rule comprehension, and multi-turn conversation memorization. Specifically, ChessArena is a competitive framework where LLMs play against each other, under four different play modes. The testbed is equipped with a ranking algorithm and a leaderboard. The testbed can also evaluate fine-grained capabilities including basic understanding, move selection, and puzzle solving. Over 13 LLMs with different modes are evaluated in ChessArena, playing over 800 games. The results reveal significant shortcomings in current LLMs: no model can beat Maia-1100 (a chess engine at human amateur level), while some even failed to defeat a random player that selects moves arbitrarily. We also present a strong baseline to the testbed: our fine-tuned Qwen3-8B substantially improved performance, approaching much larger state-of-the-art reasoning models.
Authors: Sheng Jia, Xiao Wang, Shiva Prasad Kasiviswanathan
Abstract: Although LLMs have demonstrated improved performance by scaling parallel test-time compute, doing so relies on generating reasoning paths that are both diverse and accurate. For challenging problems, the forking tokens that trigger diverse yet correct reasoning modes are typically deep in the sampling tree. Consequently, common strategies to encourage diversity, such as temperature scaling, encounter a worsened trade-off between diversity and accuracy. Motivated by this challenge, we treat parallel reasoning as a set-of-next-token-prediction problem, and incorporate a set-based global loss into Supervised Fine-Tuning (SFT) using self-supervised bipartite matching between our global forking tokens and unique reasoning traces. We observe that, while naive fine-tuning with multiple reasoning traces collapses these unique reasoning modes, our proposed method, Set Supervised Fine-Tuning (SSFT), preserves these modes and produces emergent global forking tokens. Experiments on multiple reasoning benchmarks show that our SSFT consistently outperforms SFT under both Pass@1 and Cons@k metrics.
Authors: Erjia Xiao, Lingfeng Zhang, Yingbo Tang, Hao Cheng, Renjing Xu, Wenbo Ding, Lei Zhou, Long Chen, Hangjun Ye, Xiaoshuai Hao
Abstract: In this report, we describe the technical details of our submission to the IROS 2025 RoboSense Challenge Social Navigation Track. This track focuses on developing RGBD-based perception and navigation systems that enable autonomous agents to navigate safely, efficiently, and socially compliantly in dynamic human-populated indoor environments. The challenge requires agents to operate from an egocentric perspective using only onboard sensors including RGB-D observations and odometry, without access to global maps or privileged information, while maintaining social norm compliance such as safe distances and collision avoidance. Building upon the Falcon model, we introduce a Proactive Risk Perception Module to enhance social navigation performance. Our approach augments Falcon with collision risk understanding that learns to predict distance-based collision risk scores for surrounding humans, which enables the agent to develop more robust spatial awareness and proactive collision avoidance behaviors. The evaluation on the Social-HM3D benchmark demonstrates that our method improves the agent's ability to maintain personal space compliance while navigating toward goals in crowded indoor scenes with dynamic human agents, achieving 2nd place among 16 participating teams in the challenge.
Authors: Emilio Estevan, Mar\'ia Sierra-Torralba, Eduardo L\'opez-Larraz, Luis Montesano
Abstract: Wearable EEG devices have emerged as a promising alternative to polysomnography (PSG). As affordable and scalable solutions, their widespread adoption results in the collection of massive volumes of unlabeled data that cannot be analyzed by clinicians at scale. Meanwhile, the recent success of deep learning for sleep scoring has relied on large annotated datasets. Self-supervised learning (SSL) offers an opportunity to bridge this gap, leveraging unlabeled signals to address label scarcity and reduce annotation effort. In this paper, we present the first systematic evaluation of SSL for sleep staging using wearable EEG. We investigate a range of well-established SSL methods and evaluate them on two sleep databases acquired with the Ikon Sleep wearable EEG headband: BOAS, a high-quality benchmark containing PSG and wearable EEG recordings with consensus labels, and HOGAR, a large collection of home-based, self-recorded, and unlabeled recordings. Three evaluation scenarios are defined to study label efficiency, representation quality, and cross-dataset generalization. Results show that SSL consistently improves classification performance by up to 10% over supervised baselines, with gains particularly evident when labeled data is scarce. SSL achieves clinical-grade accuracy above 80% leveraging only 5% to 10% of labeled data, while the supervised approach requires twice the labels. Additionally, SSL representations prove robust to variations in population characteristics, recording environments, and signal quality. Our findings demonstrate the potential of SSL to enable label-efficient sleep staging with wearable EEG, reducing reliance on manual annotations and advancing the development of affordable sleep monitoring systems.
Authors: Rizal Fathony, Igor Melnyk, Owen Reinert, Nam H. Nguyen, Daniele Rosa, C. Bayan Bruss
Abstract: User event modeling plays a central role in many machine learning applications, with use cases spanning e-commerce, social media, finance, cybersecurity, and other domains. User events can be broadly categorized into personal events, which involve individual actions, and relational events, which involve interactions between two users. These two types of events are typically modeled separately, using sequence-based methods for personal events and graph-based methods for relational events. Despite the need to capture both event types in real-world systems, prior work has rarely considered them together. This is often due to the convenient simplification that user behavior can be adequately represented by a single formalization, either as a sequence or a graph. To address this gap, there is a need for public datasets and prediction tasks that explicitly incorporate both personal and relational events. In this work, we introduce a collection of such datasets, propose a unified formalization, and empirically show that models benefit from incorporating both event types. Our results also indicate that current methods leave a notable room for improvements. We release these resources to support further research in unified user event modeling and encourage progress in this direction.
Authors: Hieu Le Duc, Leo Liberti
Abstract: During 2024 and 2025 the discussion about the theorem-proving capabilities of large language models started reporting interesting success stories, mostly to do with difficult exercises (such as problems from the International Mathematical Olympiad), but also with conjectures [Feldman & Karbasi, arXiv:2509.18383v1] formulated for the purpose of verifying whether the artificial intelligence could prove it. In this paper we report a theorem proving feat achieved by ChatGPT by using a protocol involving different prover and verifier instances of the gpt-5 model working collaboratively. To make sure that the produced proofs do not suffer from hallucinations, the final proof is formally verified by the lean proof assistant, and the conformance of premises and conclusion of the lean code is verified by a human. Our methodology is by no means complete or exact. It was nonetheless able to solve five out of six 2025 IMO problems, and close about a third of the sixty-six number theory conjectures in [Cohen, Journal of Integer Sequences, 2025].
Authors: Dongkwan Lee, Junhoo Lee, Nojun Kwak
Abstract: We introduce the Deep Edge Filter, a novel approach that applies high-pass filtering to deep neural network features to improve model generalizability. Our method is motivated by our hypothesis that neural networks encode task-relevant semantic information in high-frequency components while storing domain-specific biases in low-frequency components of deep features. By subtracting low-pass filtered outputs from original features, our approach isolates generalizable representations while preserving architectural integrity. Experimental results across diverse domains such as Vision, Text, 3D, and Audio demonstrate consistent performance improvements regardless of model architecture and data modality. Analysis reveals that our method induces feature sparsification and effectively isolates high-frequency components, providing empirical validation of our core hypothesis. The code is available at https://github.com/dongkwani/DeepEdgeFilter.
Authors: Guo Cheng, Danni Yang, Ziqi Huang, Jianlou Si, Chenyang Si, Ziwei Liu
Abstract: Video generative models have recently achieved notable advancements in synthesis quality. However, generating complex motions remains a critical challenge, as existing models often struggle to produce natural, smooth, and contextually consistent movements. This gap between generated and real-world motions limits their practical applicability. To address this issue, we introduce RealDPO, a novel alignment paradigm that leverages real-world data as positive samples for preference learning, enabling more accurate motion synthesis. Unlike traditional supervised fine-tuning (SFT), which offers limited corrective feedback, RealDPO employs Direct Preference Optimization (DPO) with a tailored loss function to enhance motion realism. By contrasting real-world videos with erroneous model outputs, RealDPO enables iterative self-correction, progressively refining motion quality. To support post-training in complex motion synthesis, we propose RealAction-5K, a curated dataset of high-quality videos capturing human daily activities with rich and precise motion details. Extensive experiments demonstrate that RealDPO significantly improves video quality, text alignment, and motion realism compared to state-of-the-art models and existing preference optimization techniques.
Authors: Chenwei Tang, Jingyu Xing, Xinyu Liu, Wei Ju, Jiancheng Lv, Fan Zhang, Deng Xiong, Ziyue Qiao
Abstract: Reinforcement Learning (RL) has emerged as a powerful paradigm for advancing Large Language Models (LLMs), achieving remarkable performance in complex reasoning domains such as mathematics and code generation. However, current RL methods face a fundamental scalability bottleneck due to their heavy reliance on human-curated preference data or labeled datasets for reward modeling. To overcome this limitation, we explore RL on unlabeled data where models learn autonomously from continuous experience streams. The core challenge in this setting lies in reliable reward estimation without ground-truth supervision. Existing approaches like Test-Time RL address this through self-consistent consensus, but risk reinforcing incorrect pseudo-labels derived from majority voting. We introduce COMPASS (Composite Path and Answer Self-Scoring), a novel test-time reward mechanism that operates without external supervision. COMPASS integrates two complementary components: the Dual-Calibration Answer Reward (DCAR), which stabilizes training by establishing trustworthy pseudo-labels through confidence and credibility calibration, and the Decisive Path Reward (DPR), which directly optimizes the reasoning process quality beyond mere outcome supervision. By jointly reinforcing trustworthy consensus answers and highly decisive reasoning chains, the COMPASS systematically enhances the model's analytical capabilities. Extensive experiments show that COMPASS achieves significant and consistent performance gains across diverse reasoning tasks and model architectures, advancing a more scalable direction for LLMs to learn from continuous experience.
Authors: Wang Zixian
Abstract: Direct Preference Optimization (DPO) is effective but brittle under annotator noise and distribution shift because it operates on hard, pairwise labels and only regularizes log-probability differences. We introduce Anchored Direct Preference Optimization (ADPO), a framework that extends preference learning to soft listwise supervision via reference anchoring. ADPO minimizes KL(q || softmax((s - s_ref) / tau_anc)), which (i) recovers supervised fine-tuning, knowledge distillation, maximum-entropy reinforcement learning, and DPO as special cases through suitable choices of target q, anchor policy, and temperature; (ii) induces an implicit trust region governed by the softmax Fisher metric, independent of the anchor; and (iii) supports stable dynamic-anchor updates. Empirically, we observe a task-dependent tradeoff: dynamic anchors improve online exploration under noise, while fixed anchors excel at offline distillation, achieving up to 170 to 5000 times reduction in student-teacher KL on our benchmarks.
Authors: Zhuoming Li, Aitong Liu, Mengxi Jia, Yubi Lu, Tengxiang Zhang, Changzhi Sun, Dell Zhang, Xuelong Li
Abstract: Free-form gesture understanding is highly appealing for human-computer interaction, as it liberates users from the constraints of predefined gesture categories. However, the sole existing solution GestureGPT suffers from limited recognition accuracy and slow response times. In this paper, we propose Gestura, an end-to-end system for free-form gesture understanding. Gestura harnesses a pre-trained Large Vision-Language Model (LVLM) to align the highly dynamic and diverse patterns of free-form gestures with high-level semantic concepts. To better capture subtle hand movements across different styles, we introduce a Landmark Processing Module that compensate for LVLMs' lack of fine-grained domain knowledge by embedding anatomical hand priors. Further, a Chain-of-Thought (CoT) reasoning strategy enables step-by-step semantic inference, transforming shallow knowledge into deep semantic understanding and significantly enhancing the model's ability to interpret ambiguous or unconventional gestures. Together, these components allow Gestura to achieve robust and adaptable free-form gesture comprehension. Additionally, we have developed the first open-source dataset for free-form gesture intention reasoning and understanding with over 300,000 annotated QA pairs.
Authors: Andr\'e G. Viveiros, Patrick Fernandes, Saul Santos, Sonal Sannigrahi, Emmanouil Zaranis, Nuno M. Guerreiro, Amin Farajian, Pierre Colombo, Graham Neubig, Andr\'e F. T. Martins
Abstract: Despite significant advances in vision-language models (VLMs), most existing work follows an English-centric design process, limiting their effectiveness in multilingual settings. In this work, we provide a comprehensive empirical study analyzing the impact of several multilingual design choices, such as training data composition, encoder selection, and text backbones. The result is TowerVision, a family of open multilingual VLMs for both image-text and video-text tasks, built upon the multilingual text-only model Tower+. TowerVision achieves competitive performance on multiple multimodal multilingual benchmarks and shows particular strength in culturally grounded tasks and multimodal translation. By incorporating visual and cultural context during fine-tuning, our models surpass existing approaches trained on substantially larger datasets, as demonstrated on ALM-Bench and Multi30K (image tasks) and ViMUL-Bench (video tasks). Alongside the models, we release VisionBlocks, a high-quality, curated vision-language dataset. Our findings highlight that multilingual vision-language training data substantially improves cross-lingual generalization -- both from high-resource to underrepresented languages and vice versa -- and that instruction-tuned LLMs are not always the optimal initialization point. To support further research, we publicly release all models, data, and training recipes.
Authors: Xiyu Luo, Haodong Li, Xinxing Cheng, He Zhao, Yang Hu, Xuan Song, Tianyang Zhang
Abstract: Image-to-image translation models have achieved notable success in converting images across visual domains and are increasingly used for medical tasks such as predicting post-operative outcomes and modeling disease progression. However, most existing methods primarily aim to match the target distribution and often neglect spatial correspondences between the source and translated images. This limitation can lead to structural inconsistencies and hallucinations, undermining the reliability and interpretability of the predictions. These challenges are accentuated in clinical applications by the stringent requirement for anatomical accuracy. In this work, we present TraceTrans, a novel deformable image translation model designed for post-operative prediction that generates images aligned with the target distribution while explicitly revealing spatial correspondences with the pre-operative input. The framework employs an encoder for feature extraction and dual decoders for predicting spatial deformations and synthesizing the translated image. The predicted deformation field imposes spatial constraints on the generated output, ensuring anatomical consistency with the source. Extensive experiments on medical cosmetology and brain MRI datasets demonstrate that TraceTrans delivers accurate and interpretable post-operative predictions, highlighting its potential for reliable clinical deployment.
Authors: Muhammad Faraz Ul Abrar, Nicol\`o Michelusi
Abstract: Over-the-air (OTA) federated learning (FL) has been well recognized as a scalable paradigm that exploits the waveform superposition of the wireless multiple-access channel to aggregate model updates in a single use. Existing OTA-FL designs largely enforce zero-bias model updates by either assuming \emph{homogeneous} wireless conditions (equal path loss across devices) or forcing zero-bias updates to guarantee convergence. Under \emph{heterogeneous} wireless scenarios, however, such designs are constrained by the weakest device and inflate the update variance. Moreover, prior analyses of biased OTA-FL largely address convex objectives, while most modern AI models are highly non-convex. Motivated by these gaps, we study OTA-FL with stochastic gradient descent (SGD) for general smooth non-convex objectives under wireless heterogeneity. We develop novel OTA-FL SGD updates that allow a structured, time-invariant model bias while facilitating reduced variance updates. We derive a finite-time stationarity bound (expected time average squared gradient norm) that explicitly reveals a bias-variance trade-off. To optimize this trade-off, we pose a non-convex joint OTA power-control design and develop an efficient successive convex approximation (SCA) algorithm that requires only statistical CSI at the base station. Experiments on a non-convex image classification task validate the approach: the SCA-based design accelerates convergence via an optimized bias and improves generalization over prior OTA-FL baselines.
Authors: Majid Memari, Krista Ruggles
Abstract: Artificial intelligence (AI) is transforming elementary STEM education, yet evidence remains fragmented. This systematic review synthesizes 258 studies (2020-2025) examining AI applications across eight categories: intelligent tutoring systems (45% of studies), learning analytics (18%), automated assessment (12%), computer vision (8%), educational robotics (7%), multimodal sensing (6%), AI-enhanced extended reality (XR) (4%), and adaptive content generation. The analysis shows that most studies focus on upper elementary grades (65%) and mathematics (38%), with limited cross-disciplinary STEM integration (15%). While conversational AI demonstrates moderate effectiveness (d = 0.45-0.70 where reported), only 34% of studies include standardized effect sizes. Eight major gaps limit real-world impact: fragmented ecosystems, developmental inappropriateness, infrastructure barriers, lack of privacy frameworks, weak STEM integration, equity disparities, teacher marginalization, and narrow assessment scopes. Geographic distribution is also uneven, with 90% of studies originating from North America, East Asia, and Europe. Future directions call for interoperable architectures that support authentic STEM integration, grade-appropriate design, privacy-preserving analytics, and teacher-centered implementations that enhance rather than replace human expertise.
Authors: Yuhan Cao, Yu Wang, Sitong Liu, Miao Li, Yixin Tao, Tianxing He
Abstract: The widespread adoption of Large Language Models (LLMs) through Application Programming Interfaces (APIs) induces a critical vulnerability: the potential for dishonest manipulation by service providers. This manipulation can manifest in various forms, such as secretly substituting a proclaimed high-performance model with a low-cost alternative, or inflating responses with meaningless tokens to increase billing. This work tackles the issue through the lens of algorithmic game theory and mechanism design. We are the first to propose a formal economic model for a realistic user-provider ecosystem, where a user can iteratively delegate $T$ queries to multiple model providers, and providers can engage in a range of strategic behaviors. As our central contribution, we prove that for a continuous strategy space and any $\epsilon\in(0,\frac12)$, there exists an approximate incentive-compatible mechanism with an additive approximation ratio of $O(T^{1-\epsilon}\log T)$, and a guaranteed quasi-linear second-best user utility. We also prove an impossibility result, stating that no mechanism can guarantee an expected user utility that is asymptotically better than our mechanism. Furthermore, we demonstrate the effectiveness of our mechanism in simulation experiments with real-world API settings.
Authors: Bowen Chen, Jayesh Gajbhar, Gregory Dusek, Rob Redmon, Patrick Hogan, Paul Liu, DelWayne Bohnenstiehl, Dongkuan Xu, Ruoying He
Abstract: Artificial intelligence is transforming the sciences, yet general conversational AI systems often generate unverified "hallucinations" undermining scientific rigor. We present OceanAI, a conversational platform that integrates the natural-language fluency of open-source large language models (LLMs) with real-time, parameterized access to authoritative oceanographic data streams hosted by the National Oceanic and Atmospheric Administration (NOAA). Each query such as "What was Boston Harbor's highest water level in 2024?" triggers real-time API calls that identify, parse, and synthesize relevant datasets into reproducible natural-language responses and data visualizations. In a blind comparison with three widely used AI chat-interface products, only OceanAI produced NOAA-sourced values with original data references; others either declined to answer or provided unsupported results. Designed for extensibility, OceanAI connects to multiple NOAA data products and variables, supporting applications in marine hazard forecasting, ecosystem assessment, and water-quality monitoring. By grounding outputs and verifiable observations, OceanAI advances transparency, reproducibility, and trust, offering a scalable framework for AI-enabled decision support within the oceans. A public demonstration is available at https://oceanai.ai4ocean.xyz.
Authors: Jaeyeon Lee, Hyun-Hwan Jeong, Zhandong Liu
Abstract: Diagnosing rare diseases requires linking gene findings with often unstructured reference text. Current pipelines collect many candidate genes, but clinicians still spend a lot of time filtering false positives and combining evidence from papers and databases. A key challenge is language: phenotype descriptions and inheritance patterns are written in prose, not fully captured by tables. Large language models (LLMs) can read such text, but clinical use needs grounding in citable knowledge and stable, repeatable behavior. We explore a knowledge-grounded and language-aware reranking layer on top of a high-recall first-stage pipeline. The goal is to improve precision and explainability, not to replace standard bioinformatics steps. We use expert-built context and a consensus method to reduce LLM variability, producing shorter, better-justified gene lists for expert review. LA-MARRVEL achieves the highest accuracy, outperforming other methods -- including traditional bioinformatics diagnostic tools (AI-MARRVEL, Exomiser, LIRICAL) and naive large language models (e.g., Anthropic Claude) -- with an average Recall@5 of 94.10%, a +3.65 percentage-point improvement over AI-MARRVEL. The LLM-generated reasoning provides clear prose on phenotype matching and inheritance patterns, making clinical review faster and easier. LA-MARRVEL has three parts: expert-engineered context that enriches phenotype and disease information; a ranked voting algorithm that combines multiple LLM runs to choose a consensus ranked gene list; and the AI-MARRVEL pipeline that provides first-stage ranks and gene annotations, already known as a state-of-the-art method in Rare Disease Diagnosis on BG, DDD, and UDN cohorts. The online AI-MARRVEL includes LA-MARRVEL as an LLM feature at https://ai.marrvel.org . We evaluate LA-MARRVEL on three datasets from independent cohorts of real-world diagnosed patients.
URLs: https://ai.marrvel.org
Authors: Munib Mesinovic, Max Buhlan, Tingting Zhu
Abstract: Healthcare artificial intelligence systems routinely fail when deployed across institutions, with documented performance drops and perpetuation of discriminatory patterns embedded in historical data. This brittleness stems, in part, from learning statistical associations rather than causal mechanisms. Causal graph neural networks address this triple crisis of distribution shift, discrimination, and inscrutability by combining graph-based representations of biomedical data with causal inference principles to learn invariant mechanisms rather than spurious correlations. This Review examines methodological foundations spanning structural causal models, disentangled causal representation learning, and techniques for interventional prediction and counterfactual reasoning on graphs. We analyse applications demonstrating clinical value across psychiatric diagnosis through brain network analysis, cancer subtyping via multi-omics causal integration, continuous physiological monitoring with mechanistic interpretation, and drug recommendation correcting prescription bias. These advances establish foundations for patient-specific Causal Digital Twins, enabling in silico clinical experimentation, with integration of large language models for hypothesis generation and causal graph neural networks for mechanistic validation. Substantial barriers remain, including computational requirements precluding real-time deployment, validation challenges demanding multi-modal evidence triangulation beyond cross-validation, and risks of causal-washing where methods employ causal terminology without rigorous evidentiary support. We propose tiered frameworks distinguishing causally-inspired architectures from causally-validated discoveries and identify critical research priorities making causal rather than purely associational claims.
Authors: Vivi Andersson, Sofia Bobadilla, Harald Hobbelhagen, Martin Monperrus
Abstract: Smart contracts operate in a highly adversarial environment, where vulnerabilities can lead to substantial financial losses. Thus, smart contracts are subject to security audits. In auditing, proof-of-concept (PoC) exploits play a critical role by demonstrating to the stakeholders that the reported vulnerabilities are genuine, reproducible, and actionable. However, manually creating PoCs is time-consuming, error-prone, and often constrained by tight audit schedules. We introduce POCO, an agentic framework that automatically generates executable PoC exploits from natural-language vulnerability descriptions written by auditors. POCO autonomously generates PoC exploits in an agentic manner by interacting with a set of code-execution tools in a Reason-Act-Observe loop. It produces fully executable exploits compatible with the Foundry testing framework, ready for integration into audit reports and other security tools. We evaluate POCO on a dataset of 23 real-world vulnerability reports. POCO consistently outperforms the prompting and workflow baselines, generating well-formed and logically correct PoCs. Our results demonstrate that agentic frameworks can significantly reduce the effort required for high-quality PoCs in smart contract audits. Our contribution provides readily actionable knowledge for the smart contract security community.
Authors: Jiedong Jiang, Wanyi He, Yuefeng Wang, Guoxiong Gao, Yongle Hu, Jingting Wang, Nailing Guan, Peihao Wu, Chunbo Dai, Liang Xiao, Bin Dong
Abstract: Recent advances in large language models (LLMs) have demonstrated impressive capabilities in formal theorem proving, particularly on contest-based mathematical benchmarks like the IMO. However, these contests do not reflect the depth, breadth, and abstraction of modern mathematical research. To bridge this gap, we introduce FATE (Formal Algebra Theorem Evaluation), a new benchmark series in formal algebra designed to chart a course toward advanced mathematical reasoning. We present two new components, FATE-H and FATE-X, each with 100 problems in abstract and commutative algebra. The FATE series spans a difficulty spectrum from undergraduate exercises to problems exceeding PhD qualifying exams. Notably, FATE-X is the first formal benchmark to surpass both PhD-level exam difficulty and the coverage of the Mathlib library. Our evaluations of state-of-the-art LLM provers on this new benchmark reveal a stark performance gap compared to contest math: the best model achieves only 3% (pass@64) accuracy on FATE-H and 0% on FATE-X. Our two-stage evaluation reveals that models' natural-language reasoning is notably more accurate than their ability to formalize this reasoning. We systematically classify the common errors that arise during this formalization process. Furthermore, a comparative study shows that a specialized prover can exhibit less effective reflection than general-purpose models, reducing its accuracy at the natural-language stage. We believe FATE provides a robust and challenging benchmark that establishes essential checkpoints on the path toward research-level formal mathematical reasoning.
Authors: Gian Maria Campedelli
Abstract: While the possibility of reaching human-like Artificial Intelligence (AI) remains controversial, the likelihood that the future will be characterized by a society with a growing presence of autonomous machines is high. Autonomous AI agents are already deployed and active across several industries and digital environments and alongside human-human and human-machine interactions, machine-machine interactions are poised to become increasingly prevalent. Given these developments, I argue that criminology must begin to address the implications of this transition for crime and social control. Drawing on Actor-Network Theory and Woolgar's decades-old call for a sociology of machines -- frameworks that acquire renewed relevance with the rise of generative AI agents -- I contend that criminologists should move beyond conceiving AI solely as a tool. Instead, AI agents should be recognized as entities with agency encompassing computational, social, and legal dimensions. Building on the literature on AI safety, I thus examine the risks associated with the rise of multi-agent AI systems, proposing a dual taxonomy to characterize the channels through which interactions among AI agents may generate deviant, unlawful, or criminal outcomes. I then advance and discuss four key questions that warrant theoretical and empirical attention: (1) Can we assume that machines will simply mimic humans? (2) Will crime theories developed for humans suffice to explain deviant or criminal behaviors emerging from interactions between autonomous AI agents? (3) What types of criminal behaviors will be affected first? (4) How might this unprecedented societal shift impact policing? These questions underscore the urgent need for criminologists to theoretically and empirically engage with the implications of multi-agent AI systems for the study of crime and play a more active role in debates on AI safety and governance.
Authors: Yuya Miyaoka, Masaki Inoue
Abstract: This paper proposes a control-based framework for aligning large language models (LLMs) by leveraging a control barrier function (CBF) to ensure user-desirable text generation. The presented framework applies the CBF safety filter to the predicted token generated from the baseline LLM, to intervene in the generated text. The safety filter includes two significant advantages: this safety filter is an add-on type, allowing it to be used for alignment purposes without fine-tuning the baseline LLM, and if there is an evaluation model regarding the desired alignment, it can be directly applied to the filter design. The overall text-generation system is implemented with open-source language models, aiming to generate positive text.
Authors: Alexander Htet Kyaw, Lenin Ravindranath Sivalingam
Abstract: We present a node-based storytelling system for multimodal content generation. The system represents stories as graphs of nodes that can be expanded, edited, and iteratively refined through direct user edits and natural-language prompts. Each node can integrate text, images, audio, and video, allowing creators to compose multimodal narratives. A task selection agent routes between specialized generative tasks that handle story generation, node structure reasoning, node diagram formatting, and context generation. The interface supports targeted editing of individual nodes, automatic branching for parallel storylines, and node-based iterative refinement. Our results demonstrate that node-based editing supports control over narrative structure and iterative generation of text, images, audio, and video. We report quantitative outcomes on automatic story outline generation and qualitative observations of editing workflows. Finally, we discuss current limitations such as scalability to longer narratives and consistency across multiple nodes, and outline future work toward human-in-the-loop and user-centered creative AI tools.
Authors: Doria Bonzi, Alexandre Guiggi, Fr\'ed\'eric B\'echet, Carlos Ramisch, Benoit Favre
Abstract: Critical appraisal of scientific literature is an essential skill in the biomedical field. While large language models (LLMs) can offer promising support in this task, their reliability remains limited, particularly for critical reasoning in specialized domains. We introduce CareMedEval, an original dataset designed to evaluate LLMs on biomedical critical appraisal and reasoning tasks. Derived from authentic exams taken by French medical students, the dataset contains 534 questions based on 37 scientific articles. Unlike existing benchmarks, CareMedEval explicitly evaluates critical reading and reasoning grounded in scientific papers. Benchmarking state-of-the-art generalist and biomedical-specialized LLMs under various context conditions reveals the difficulty of the task: open and commercial models fail to exceed an Exact Match Rate of 0.5 even though generating intermediate reasoning tokens considerably improves the results. Yet, models remain challenged especially on questions about study limitations and statistical analysis. CareMedEval provides a challenging benchmark for grounded reasoning, exposing current LLM limitations and paving the way for future development of automated support for critical appraisal.
Authors: Gabriel Spadon, Vaishnav Vaidheeswaran, Claudio DiBacco
Abstract: Marine invasive species spread through global shipping and generate substantial ecological and economic impacts. Traditional risk assessments require detailed records of ballast water and traffic patterns, which are often incomplete, limiting global coverage. This work advances a theoretical framework that quantifies invasion risk by combining environmental similarity across ports with observed and forecasted maritime mobility. Climate-based feature representations characterize each port's marine conditions, while mobility networks derived from Automatic Identification System data capture vessel flows and potential transfer pathways. Clustering and metric learning reveal climate analogues and enable the estimation of species survival likelihood along shipping routes. A temporal link prediction model captures how traffic patterns may change under shifting environmental conditions. The resulting fusion of environmental similarity and predicted mobility provides exposure estimates at the port and voyage levels, supporting targeted monitoring, routing adjustments, and management interventions.