Authors: Erdi Gao, Haowei Yang, Dan Sun, Haohao Xia, Yuhan Ma, Yuanjing Zhu
Abstract: In the field of natural language processing, text classification, as a basic task, has important research value and application prospects. Traditional text classification methods usually rely on feature representations such as the bag of words model or TF-IDF, which overlook the semantic connections between words and make it challenging to grasp the deep structural details of the text. Recently, GNNs have proven to be a valuable asset for text classification tasks, thanks to their capability to handle non-Euclidean data efficiently. However, the existing text classification methods based on GNN still face challenges such as complex graph structure construction and high cost of model training. This paper introduces a text classification optimization algorithm utilizing graph neural networks. By introducing adaptive graph construction strategy and efficient graph convolution operation, the accuracy and efficiency of text classification are effectively improved. The experimental results demonstrate that the proposed method surpasses traditional approaches and existing GNN models across multiple public datasets, highlighting its superior performance and feasibility for text classification tasks.
Authors: Christopher Sun, Abishek Satish
Abstract: In this study, we implement a novel BERT architecture for multitask fine-tuning on three downstream tasks: sentiment classification, paraphrase detection, and semantic textual similarity prediction. Our model, Multitask BERT, incorporates layer sharing and a triplet architecture, custom sentence pair tokenization, loss pairing, and gradient surgery. Such optimizations yield a 0.516 sentiment classification accuracy, 0.886 paraphase detection accuracy, and 0.864 semantic textual similarity correlation on test data. We also apply generative adversarial learning to BERT, constructing a conditional generator model that maps from latent space to create fake embeddings in $\mathbb{R}^{768}$. These fake embeddings are concatenated with real BERT embeddings and passed into a discriminator model for auxiliary classification. Using this framework, which we refer to as AC-GAN-BERT, we conduct semi-supervised sensitivity analyses to investigate the effect of increasing amounts of unlabeled training data on AC-GAN-BERT's test accuracy. Overall, aside from implementing a high-performing multitask classification system, our novelty lies in the application of adversarial learning to construct a generator that mimics BERT. We find that the conditional generator successfully produces rich embeddings with clear spatial correlation with class labels, demonstrating avoidance of mode collapse. Our findings validate the GAN-BERT approach and point to future directions of generator-aided knowledge distillation.
Authors: Vil\'em Zouhar, Pinzhen Chen, Tsz Kin Lam, Nikita Moghe, Barry Haddow
Abstract: Since its introduction, the COMET metric has blazed a trail in the machine translation community, given its strong correlation with human judgements of translation quality. Its success stems from being a modified pre-trained multilingual model finetuned for quality assessment. However, it being a machine learning model also gives rise to a new set of pitfalls that may not be widely known. We investigate these unexpected behaviours from three aspects: 1) technical: obsolete software versions and compute precision; 2) data: empty content, language mismatch, and translationese at test time as well as distribution and domain biases in training; 3) usage and reporting: multi-reference support and model referencing in the literature. All of these problems imply that COMET scores is not comparable between papers or even technical setups and we put forward our perspective on fixing each issue. Furthermore, we release the SacreCOMET package that can generate a signature for the software and model configuration as well as an appropriate citation. The goal of this work is to help the community make more sound use of the COMET metric.
Authors: Adamu Lawan, Juhua Pu, Haruna Yunusa, Muhammad Lawan, Aliyu Umar, Adamu Sani Yahya
Abstract: Multimodal aspect-based sentiment analysis (MABSA) enhances sentiment detection by combining text with other data types like images. However, despite setting significant benchmarks, attention mechanisms exhibit limitations in efficiently modelling long-range dependencies between aspect and opinion targets within the text. They also face challenges in capturing global-context dependencies for visual representations. To this end, we propose Kolmogorov-Arnold Networks (KANs) and Selective State Space model (Mamba) transformer (DualKanbaFormer), a novel architecture to address the above issues. We leverage the power of Mamba to capture global context dependencies, Multi-head Attention (MHA) to capture local context dependencies, and KANs to capture non-linear modelling patterns for both textual representations (textual KanbaFormer) and visual representations (visual KanbaFormer). Furthermore, we fuse the textual KanbaFormer and visual KanbaFomer with a gated fusion layer to capture the inter-modality dynamics. According to extensive experimental results, our model outperforms some state-of-the-art (SOTA) studies on two public datasets.
Authors: Adrian de Wynter
Abstract: We perform a critical examination of the scientific methodology behind contemporary large language model (LLM) research. For this we assess over 2,000 research works based on criteria typical of what is considered good research (e.g. presence of statistical tests and reproducibility) and cross-validate it with arguments that are at the centre of controversy (e.g., claims of emergent behaviour, the use of LLMs as evaluators). We find multiple trends, such as declines in claims of emergent behaviour and the presence of ethics disclaimers; and the rise of LLMs as evaluators. This paper underscores the need for more scrutiny and rigour by and from this field. Critical reading and familiarity with the literature are crucial to live up to the fundamentals of a responsible scientific method that is ethical, reproducible, systematic, and open to criticism.
Authors: Yize Zhao, Tina Behnia, Vala Vakilian, Christos Thrampoulidis
Abstract: Next-token prediction (NTP) over large text corpora has become the go-to paradigm to train large language models. Yet, it remains unclear how NTP influences the mapping of linguistic patterns to geometric properties of the resulting model representations. We frame training of large language models as soft-label classification over sparse probabilistic label vectors, coupled with an analytical approximation that allows unrestricted generation of context embeddings. This approach links NTP training to rank-constrained, nuclear-norm regularized optimization in the logit domain, offering a framework for analyzing the geometry of word and context embeddings. In large embedding spaces, we find that NTP implicitly favors learning logits with a sparse plus low-rank structure. While the sparse component captures the co-occurrence frequency of context-word pairs, the orthogonal low-rank component, which becomes dominant as training progresses, depends solely on the sparsity pattern of the co-occurrence matrix. Consequently, when projected onto an appropriate subspace, representations of contexts that are followed by the same set of next-tokens collapse, a phenomenon we term subspace-collapse. We validate our findings on synthetic and small-scale real language datasets. Finally, we outline potential research directions aimed at deepening the understanding of NTP's influence on the learning of linguistic patterns and regularities.
Authors: Jialin Wu, Jiangyi Deng, Shengyuan Pang, Yanjiao Chen, Jiayang Xu, Xinfeng Li, Wenyuan Xu
Abstract: Given the societal impact of unsafe content generated by large language models (LLMs), ensuring that LLM services comply with safety standards is a crucial concern for LLM service providers. Common content moderation methods are limited by an effectiveness-and-efficiency dilemma, where simple models are fragile while sophisticated models consume excessive computational resources. In this paper, we reveal for the first time that effective and efficient content moderation can be achieved by extracting conceptual features from chat-oriented LLMs, despite their initial fine-tuning for conversation rather than content moderation. We propose a practical and unified content moderation framework for LLM services, named Legilimens, which features both effectiveness and efficiency. Our red-team model-based data augmentation enhances the robustness of Legilimens against state-of-the-art jailbreaking. Additionally, we develop a framework to theoretically analyze the cost-effectiveness of Legilimens compared to other methods. We have conducted extensive experiments on five host LLMs, seventeen datasets, and nine jailbreaking methods to verify the effectiveness, efficiency, and robustness of Legilimens against normal and adaptive adversaries. A comparison of Legilimens with both commercial and academic baselines demonstrates the superior performance of Legilimens. Furthermore, we confirm that Legilimens can be applied to few-shot scenarios and extended to multi-label classification tasks.
Authors: Haowen Hou, Fei Ma, Binwen Bai, Xinxin Zhu, Fei Yu
Abstract: Large Language Models (LLMs) have garnered widespread attention due to their remarkable performance across various tasks. However, to mitigate the issue of hallucinations, LLMs often incorporate retrieval-augmented pipeline to provide them with rich external knowledge and context. Nevertheless, challenges stem from inaccurate and coarse-grained context retrieved from the retriever. Supplying irrelevant context to the LLMs can result in poorer responses, increased inference latency, and higher costs. This paper introduces a method called Instruction-Aware Contextual Compression, which filters out less informative content, thereby accelerating and enhancing the use of LLMs. The experimental results demonstrate that Instruction-Aware Contextual Compression notably reduces memory consumption and minimizes generation latency while maintaining performance levels comparable to those achieved with the use of the full context. Specifically, we achieved a 50% reduction in context-related costs, resulting in a 5% reduction in inference memory usage and a 2.2-fold increase in inference speed, with only a minor drop of 0.047 in Rouge-1. These findings suggest that our method strikes an effective balance between efficiency and performance.
Authors: Danlong Yuan, Jiahao Liu, Bei Li, Huishuai Zhang, Jingang Wang, Xunliang Cai, Dongyan Zhao
Abstract: While the Mamba architecture demonstrates superior inference efficiency and competitive performance on short-context natural language processing (NLP) tasks, empirical evidence suggests its capacity to comprehend long contexts is limited compared to transformer-based models. In this study, we investigate the long-context efficiency issues of the Mamba models and propose ReMamba, which enhances Mamba's ability to comprehend long contexts. ReMamba incorporates selective compression and adaptation techniques within a two-stage re-forward process, incurring minimal additional inference costs overhead. Experimental results on the LongBench and L-Eval benchmarks demonstrate ReMamba's efficacy, improving over the baselines by 3.2 and 1.6 points, respectively, and attaining performance almost on par with same-size transformer models.
Authors: Wei Chen, Zhiyuan Li, Shuo Xin, Yihao Wang
Abstract: This paper presents Dolphin, a novel decoder-decoder architecture for energy-efficient processing of long contexts in language models. Our approach addresses the significant energy consumption and latency challenges inherent in on-device models. Dolphin employs a compact 0.5B parameter decoder to distill extensive contextual information into a memory embedding, substantially reducing the input length for the primary 7B parameter decoder model. Inspired by vision-language models, we repurpose the image embedding projector to encode long textual contexts, effectively treating extended context as a distinct modality. This innovative method enables processing of substantially longer contexts without the typical computational overhead associated with extended input sequences. Empirical evaluations demonstrate a 10-fold improvement in energy efficiency and a 5-fold reduction in latency compared to conventional full-length context processing methods without losing quality of the response. Our work contributes to the development of more sustainable and scalable language models for on-device applications, addressing the critical need for energy-efficient and responsive AI technologies in resource-constrained environments while maintaining the accuracy to understand long contexts. This research has implications for the broader field of natural language processing, particularly in the domain of efficient model design for resource-limited settings. By enabling more sophisticated AI capabilities on edge devices, Dolphin paves the way for advanced language processing in a wide range of applications where computational resources are at a premium. The Dolphin model is publicly available at https://huggingface.co/NexaAIDev/Dolphin.
Authors: Haichuan Hu, Yuhan Sun, Qunjun Zhang
Abstract: Retrieval-Augmented Generation (RAG) has become a primary technique for mitigating hallucinations in large language models (LLMs). However, incomplete knowledge extraction and insufficient understanding can still mislead LLMs to produce irrelevant or even contradictory responses, which means hallucinations persist in RAG. In this paper, we propose LRP4RAG, a method based on the Layer-wise Relevance Propagation (LRP) algorithm for detecting hallucinations in RAG. Specifically, we first utilize LRP to compute the relevance between the input and output of the RAG generator. We then apply further extraction and resampling to the relevance matrix. The processed relevance data are input into multiple classifiers to determine whether the output contains hallucinations. To the best of our knowledge, this is the first time that LRP has been used for detecting RAG hallucinations, and extensive experiments demonstrate that LRP4RAG outperforms existing baselines.
Authors: Yunmeng Li, Jun Suzuki, Makoto Morishita, Kaori Abe, Kentaro Inui
Abstract: The complexities of chats pose significant challenges for machine translation models. Recognizing the need for a precise evaluation metric to address the issues of chat translation, this study introduces Multidimensional Quality Metrics for Chat Translation (MQM-Chat). Through the experiments of five models using MQM-Chat, we observed that all models generated certain fundamental errors, while each of them has different shortcomings, such as omission, overly correcting ambiguous source content, and buzzword issues, resulting in the loss of stylized information. Our findings underscore the effectiveness of MQM-Chat in evaluating chat translation, emphasizing the importance of stylized content and dialogue consistency for future studies.
Authors: Taiwei Shi, Zhuoer Wang, Longqi Yang, Ying-Chun Lin, Zexue He, Mengting Wan, Pei Zhou, Sujay Jauhar, Xiaofeng Xu, Xia Song, Jennifer Neville
Abstract: As large language models (LLMs) continue to advance, aligning these models with human preferences has emerged as a critical challenge. Traditional alignment methods, relying on human or LLM annotated datasets, are limited by their resource-intensive nature, inherent subjectivity, and the risk of feedback loops that amplify model biases. To overcome these limitations, we introduce WildFeedback, a novel framework that leverages real-time, in-situ user interactions to create preference datasets that more accurately reflect authentic human values. WildFeedback operates through a three-step process: feedback signal identification, preference data construction, and user-guided evaluation. We applied this framework to a large corpus of user-LLM conversations, resulting in a rich preference dataset that reflects genuine user preferences. This dataset captures the nuances of user preferences by identifying and classifying feedback signals within natural conversations, thereby enabling the construction of more representative and context-sensitive alignment data. Our extensive experiments demonstrate that LLMs fine-tuned on WildFeedback exhibit significantly improved alignment with user preferences, as evidenced by both traditional benchmarks and our proposed user-guided evaluation. By incorporating real-time feedback from actual users, WildFeedback addresses the scalability, subjectivity, and bias challenges that plague existing approaches, marking a significant step toward developing LLMs that are more responsive to the diverse and evolving needs of their users. In summary, WildFeedback offers a robust, scalable solution for aligning LLMs with true human values, setting a new standard for the development and evaluation of user-centric language models.
Authors: Lujun Gui, Bin Xiao, Lei Su, Weipeng Chen
Abstract: Lossless speculative decoding accelerates target large language model (LLM) inference by employing a lightweight draft model for generating tree-structured candidates, which are subsequently verified in parallel by the target LLM. Currently, effective approaches leverage feature-level rather than token-level autoregression within the draft model to facilitate more straightforward predictions and enhanced knowledge distillation. In this paper, we reassess these approaches and propose FSPAD (Feature Sampling and Partial Alignment Distillation for Lossless Speculative Decoding), which introduces two straightforward and effective components within the existing framework to boost lossless speculative decoding. Firstly, FSPAD utilizes token embeddings to sample features of the target LLM in high-dimensional space before feeding them into the draft model, due to the inherent uncertainty of the features preventing the draft model from obtaining the specific token output by the target LLM. Secondly, FSPAD introduces partial alignment distillation to weaken the draft model's connection between features and logits, aiming to reduce the conflict between feature alignment and logit confidence during training. Our experiments include both greedy and non-greedy decoding on the largest and smallest models from the Vicuna and LLaMA3-Instruct series, as well as tasks in multi-turn conversation, translation, summarization, question answering, mathematical reasoning, and retrieval-augmented generation. The results show that FSPAD outperforms the state-of-the-art method across all the aforementioned tasks and target LLMs.
Authors: Dian Yu, Baolin Peng, Ye Tian, Linfeng Song, Haitao Mi, Dong Yu
Abstract: There is a growing trend of teaching large language models (LLMs) to solve mathematical problems through coding. Existing studies primarily focus on prompting powerful, closed-source models to generate seed training data followed by in-domain data augmentation, equipping LLMs with considerable capabilities for code-aided mathematical reasoning. However, continually training these models on augmented data derived from a few datasets such as GSM8K may impair their generalization abilities and restrict their effectiveness to a narrow range of question types. Conversely, the potential of improving such LLMs by leveraging large-scale, expert-written, diverse math question-answer pairs remains unexplored. To utilize these resources and tackle unique challenges such as code response assessment, we propose a novel paradigm that uses a code-based critic model to guide steps including question-code data construction, quality control, and complementary evaluation. We also explore different alignment algorithms with self-generated instruction/preference data to foster continuous improvement. Experiments across both in-domain (up to +5.7%) and out-of-domain (+4.4%) benchmarks in English and Chinese demonstrate the effectiveness of the proposed paradigm.
Authors: Korbinian Kuhn, Verena Kersken, Gottfried Zimmermann
Abstract: The Word Error Rate (WER) is the common measure of accuracy for Automatic Speech Recognition (ASR). Transcripts are usually pre-processed by substituting specific characters to account for non-semantic differences. As a result of this normalisation, information on the accuracy of punctuation or capitalisation is lost. We present a non-destructive, token-based approach using an extended Levenshtein distance algorithm to compute a robust WER and additional orthographic metrics. Transcription errors are also classified more granularly by existing string similarity and phonetic algorithms. An evaluation on several datasets demonstrates the practical equivalence of our approach compared to common WER computations. We also provide an exemplary analysis of derived use cases, such as a punctuation error rate, and a web application for interactive use and visualisation of our implementation. The code is available open-source.
Authors: Lingyu Gao
Abstract: Text classification is crucial for applications such as sentiment analysis and toxic text filtering, but it still faces challenges due to the complexity and ambiguity of natural language. Recent advancements in deep learning, particularly transformer architectures and large-scale pretraining, have achieved inspiring success in NLP fields. Building on these advancements, this thesis explores three challenging settings in text classification by leveraging the intrinsic knowledge of pretrained language models (PLMs). Firstly, to address the challenge of selecting misleading yet incorrect distractors for cloze questions, we develop models that utilize features based on contextualized word representations from PLMs, achieving performance that rivals or surpasses human accuracy. Secondly, to enhance model generalization to unseen labels, we create small finetuning datasets with domain-independent task label descriptions, improving model performance and robustness. Lastly, we tackle the sensitivity of large language models to in-context learning prompts by selecting effective demonstrations, focusing on misclassified examples and resolving model ambiguity regarding test example labels.
Authors: Jillian Fisher, Skyler Hallinan, Ximing Lu, Mitchell Gordon, Zaid Harchaoui, Yejin Choi
Abstract: Authorship obfuscation, rewriting a text to intentionally obscure the identity of the author, is an important but challenging task. Current methods using large language models (LLMs) lack interpretability and controllability, often ignoring author-specific stylistic features, resulting in less robust performance overall. To address this, we develop StyleRemix, an adaptive and interpretable obfuscation method that perturbs specific, fine-grained style elements of the original input text. StyleRemix uses pre-trained Low Rank Adaptation (LoRA) modules to rewrite an input specifically along various stylistic axes (e.g., formality and length) while maintaining low computational cost. StyleRemix outperforms state-of-the-art baselines and much larger LLMs in a variety of domains as assessed by both automatic and human evaluation. Additionally, we release AuthorMix, a large set of 30K high-quality, long-form texts from a diverse set of 14 authors and 4 domains, and DiSC, a parallel corpus of 1,500 texts spanning seven style axes in 16 unique directions
Authors: Talia Tseriotou, Adam Tsakalidis, Maria Liakata
Abstract: Dynamic representation learning plays a pivotal role in understanding the evolution of linguistic content over time. On this front both context and time dynamics as well as their interplay are of prime importance. Current approaches model context via pre-trained representations, which are typically temporally agnostic. Previous work on modeling context and temporal dynamics has used recurrent methods, which are slow and prone to overfitting. Here we introduce TempoFormer, the fist task-agnostic transformer-based and temporally-aware model for dynamic representation learning. Our approach is jointly trained on inter and intra context dynamics and introduces a novel temporal variation of rotary positional embeddings. The architecture is flexible and can be used as the temporal representation foundation of other models or applied to different transformer-based architectures. We show new SOTA performance on three different real-time change detection tasks.
Authors: Shiyu Li, Yang Tang, Shizhe Chen, Xi Chen
Abstract: With the growing popularity of RAG, the capabilities of embedding models are gaining increasing attention. Embedding models are primarily trained through contrastive loss learning, with negative examples being a key component. Previous work has proposed various hard negative mining strategies, but these strategies are typically employed as preprocessing steps. In this paper, we propose the conan-embedding model, which maximizes the utilization of more and higher-quality negative examples. Specifically, since the model's ability to handle preprocessed negative examples evolves during training, we propose dynamic hard negative mining method to expose the model to more challenging negative examples throughout the training process. Secondly, contrastive learning requires as many negative examples as possible but is limited by GPU memory constraints. Therefore, we use a Cross-GPU balancing Loss to provide more negative examples for embedding training and balance the batch size across multiple tasks. Moreover, we also discovered that the prompt-response pairs from LLMs can be used for embedding training. Our approach effectively enhances the capabilities of embedding models, currently ranking first on the Chinese leaderboard of Massive text embedding benchmark
Authors: Wazir Ali, Saifullah Tumrani, Jay Kumar, Tariq Rahim Soomro
Abstract: In this paper, we propose a new word embedding based corpus consisting of more than 61 million words crawled from multiple web resources. We design a preprocessing pipeline for the filtration of unwanted text from crawled data. Afterwards, the cleaned vocabulary is fed to state-of-the-art continuous-bag-of-words, skip-gram, and GloVe word embedding algorithms. For the evaluation of pretrained embeddings, we use popular intrinsic and extrinsic evaluation approaches. The evaluation results reveal that continuous-bag-of-words and skip-gram perform better than GloVe and existing Sindhi fastText word embedding on both intrinsic and extrinsic evaluation approaches
Authors: Max Ploner, Jacek Wiland, Sebastian Pohl, Alan Akbik
Abstract: Knowledge probing evaluates the extent to which a language model (LM) has acquired relational knowledge during its pre-training phase. It provides a cost-effective means of comparing LMs of different sizes and training setups and is useful for monitoring knowledge gained or lost during continual learning (CL). In prior work, we presented an improved knowledge probe called BEAR (Wiland et al., 2024), which enables the comparison of LMs trained with different pre-training objectives (causal and masked LMs) and addresses issues of skewed distributions in previous probes to deliver a more unbiased reading of LM knowledge. With this paper, we present LM-PUB- QUIZ, a Python framework and leaderboard built around the BEAR probing mechanism that enables researchers and practitioners to apply it in their work. It provides options for standalone evaluation and direct integration into the widely-used training pipeline of the Hugging Face TRANSFORMERS library. Further, it provides a fine-grained analysis of different knowledge types to assist users in better understanding the knowledge in each evaluated LM. We publicly release LM-PUB-QUIZ as an open-source project.
Authors: Yuxin Lu, Yu-Ying Chuang, R. Harald Baayen
Abstract: In Standard Chinese, Tone 3 (the dipping tone) becomes Tone 2 (rising tone) when followed by another Tone 3. Previous studies have noted that this sandhi process may be incomplete, in the sense that the assimilated Tone 3 is still distinct from a true Tone 2. While Mandarin Tone 3 sandhi is widely studied using carefully controlled laboratory speech (Xu, 1997) and more formal registers of Beijing Mandarin (Yuan and Chen, 2014), less is known about its realization in spontaneous speech, and about the effect of contextual factors on tonal realization. The present study investigates the pitch contours of two-character words with T2-T3 and T3-T3 tone patterns in spontaneous Taiwan Mandarin conversations. Our analysis makes use of the Generative Additive Mixed Model (GAMM, Wood, 2017) to examine fundamental frequency (f0) contours as a function of normalized time. We consider various factors known to influence pitch contours, including gender, speaking rate, speaker, neighboring tones, word position, bigram probability, and also novel predictors, word and word sense (Chuang et al., 2024). Our analyses revealed that in spontaneous Taiwan Mandarin, T3-T3 words become indistinguishable from T2-T3 words, indicating complete sandhi, once the strong effect of word (or word sense) is taken into account. For our data, the shape of f0 contours is not co-determined by word frequency. In contrast, the effect of word meaning on f0 contours is robust, as strong as the effect of adjacent tones, and is present for both T2-T3 and T3-T3 words.
Authors: Huachuan Qiu, Zhenzhong Lan
Abstract: Virtual counselors powered by large language models (LLMs) aim to create interactive support systems that effectively assist clients struggling with mental health challenges. To replicate counselor-client conversations, researchers have built an online mental health platform that allows professional counselors to provide clients with text-based counseling services for about an hour per session. Notwithstanding its effectiveness, challenges exist as human annotation is time-consuming, cost-intensive, privacy-protected, and not scalable. To address this issue and investigate the applicability of LLMs in psychological counseling conversation simulation, we propose a framework that employs two LLMs via role-playing for simulating counselor-client interactions. Our framework involves two LLMs, one acting as a client equipped with a specific and real-life user profile and the other playing the role of an experienced counselor, generating professional responses using integrative therapy techniques. We implement both the counselor and the client by zero-shot prompting the GPT-4 model. In order to assess the effectiveness of LLMs in simulating counselor-client interactions and understand the disparities between LLM- and human-generated conversations, we evaluate the synthetic data from various perspectives. We begin by assessing the client's performance through automatic evaluations. Next, we analyze and compare the disparities between dialogues generated by the LLM and those generated by professional counselors. Furthermore, we conduct extensive experiments to thoroughly examine the performance of our LLM-based counselor trained with synthetic interactive dialogues by benchmarking against state-of-the-art models for mental health.
Authors: Konstantin Dobler, Gerard de Melo
Abstract: We investigate continued pretraining of LLMs for language adaptation on a tight academic budget: a setting in which only a few GPUs can be used in parallel, for a heavily constrained duration. We focus on adapting Mistral-7B to German or Arabic and evaluate several techniques to improve efficiency and effectiveness in this setting. Our German models adapted on this tight compute budget underperform compared to the base Mistral-7B, while our Arabic models outperform several baselines, showing that for sufficiently well-represented languages, continued pretraining for specialization is not always helpful. Our main findings focus on training precision and tokenizer swapping. Our results show that pure bfloat16 training is a viable alternative to mixed-precision training, while being much faster when only using a few GPUs. Swapping the tokenizer for a specialized one yields more efficient tokenization and is competitive with the original tokenizer, which already contains some German tokens, but did not significantly increase performance for German. Code and model weights are available at on GitHub.
Authors: L\'eo Hemamou, Mehdi Debiane
Abstract: In an era where digital text is proliferating at an unprecedented rate, efficient summarization tools are becoming indispensable. While Large Language Models (LLMs) have been successfully applied in various NLP tasks, their role in extractive text summarization remains underexplored. This paper introduces EYEGLAXS (Easy Yet Efficient larGe LAnguage model for eXtractive Summarization), a framework that leverages LLMs, specifically LLAMA2-7B and ChatGLM2-6B, for extractive summarization of lengthy text documents. Instead of abstractive methods, which often suffer from issues like factual inaccuracies and hallucinations, EYEGLAXS focuses on extractive summarization to ensure factual and grammatical integrity. Utilizing state-of-the-art techniques such as Flash Attention and Parameter-Efficient Fine-Tuning (PEFT), EYEGLAXS addresses the computational and resource challenges typically associated with LLMs. The system sets new performance benchmarks on well-known datasets like PubMed and ArXiv. Furthermore, we extend our research through additional analyses that explore the adaptability of LLMs in handling different sequence lengths and their efficiency in training on smaller datasets. These contributions not only set a new standard in the field but also open up promising avenues for future research in extractive text summarization.
Authors: Sebastian Vallejo Vera, Hunter Driggers
Abstract: Human coders are biased. We test similar biases in Large Language Models (LLMs) as annotators. By replicating an experiment run by Ennser-Jedenastik and Meyer (2018), we find evidence that LLMs use political information, and specifically party cues, to judge political statements. Not only do LLMs use relevant information to contextualize whether a statement is positive, negative, or neutral based on the party cue, they also reflect the biases of the human-generated data upon which they have been trained. We also find that unlike humans, who are only biased when faced with statements from extreme parties, LLMs exhibit significant bias even when prompted with statements from center-left and center-right parties. The implications of our findings are discussed in the conclusion.
Authors: Mohammad Ebrahimi, Behrouz Minaei Bidgoli, Nasim Khozouei
Abstract: Semantic role labeling is a crucial task in natural language processing, enabling better comprehension of natural language. However, the lack of annotated data in multiple languages has posed a challenge for researchers. To address this, a deep learning algorithm based on model transfer has been proposed. The algorithm utilizes a dataset consisting of the English portion of CoNLL2009 and a corpus of semantic roles in Persian. To optimize the efficiency of training, only ten percent of the educational data from each language is used. The results of the proposed model demonstrate significant improvements compared to Niksirt et al.'s model. In monolingual mode, the proposed model achieved a 2.05 percent improvement on F1-score, while in cross-lingual mode, the improvement was even more substantial, reaching 6.23 percent. Worth noting is that the compared model only trained two of the four stages of semantic role labeling and employed golden data for the remaining two stages. This suggests that the actual superiority of the proposed model surpasses the reported numbers by a significant margin. The development of cross-lingual methods for semantic role labeling holds promise, particularly in addressing the scarcity of annotated data for various languages. These advancements pave the way for further research in understanding and processing natural language across different linguistic contexts.
Authors: Nikolas Gritsch, Qizhen Zhang, Acyr Locatelli, Sara Hooker, Ahmet \"Ust\"un
Abstract: Efficiency, specialization, and adaptability to new data distributions are qualities that are hard to combine in current Large Language Models. The Mixture of Experts (MoE) architecture has been the focus of significant research because its inherent conditional computation enables such desirable properties. In this work, we focus on "upcycling" dense expert models into an MoE, aiming to improve specialization while also adding the ability to adapt to new tasks easily. We introduce Nexus, an enhanced MoE architecture with adaptive routing where the model learns to project expert embeddings from domain representations. This approach allows Nexus to flexibly add new experts after the initial upcycling through separately trained dense models, without requiring large-scale MoE training for unseen data domains. Our experiments show that Nexus achieves a relative gain of up to 2.1% over the baseline for initial upcycling, and a 18.8% relative gain for extending the MoE with a new expert by using limited finetuning data. This flexibility of Nexus is crucial to enable an open-source ecosystem where every user continuously assembles their own MoE-mix according to their needs.
Authors: Ruirui Chen, Weifeng Jiang, Chengwei Qin, Ishaan Singh Rawal, Cheston Tan, Dongkyu Choi, Bo Xiong, Bo Ai
Abstract: The rapid obsolescence of information in Large Language Models (LLMs) has driven the development of various techniques to incorporate new facts. However, existing methods for knowledge editing still face difficulties with multi-hop questions that require accurate fact identification and sequential logical reasoning, particularly among numerous fact updates. To tackle these challenges, this paper introduces Graph Memory-based Editing for Large Language Models (GMeLLo), a straitforward and effective method that merges the explicit knowledge representation of Knowledge Graphs (KGs) with the linguistic flexibility of LLMs. Beyond merely leveraging LLMs for question answering, GMeLLo employs these models to convert free-form language into structured queries and fact triples, facilitating seamless interaction with KGs for rapid updates and precise multi-hop reasoning. Our results show that GMeLLo significantly surpasses current state-of-the-art knowledge editing methods in the multi-hop question answering benchmark, MQuAKE, especially in scenarios with extensive knowledge edits.
Authors: Wei Wang, Dan Zhang, Tao Feng, Boyan Wang, Jie Tang
Abstract: Large Language Models (LLMs) are becoming increasingly powerful and capable of handling complex tasks, e.g., building single agents and multi-agent systems. Compared to single agents, multi-agent systems have higher requirements for the collaboration capabilities of language models. Many benchmarks are proposed to evaluate their collaborative abilities. However, these benchmarks lack fine-grained evaluations of LLM collaborative capabilities. Additionally, multi-agent collaborative and competitive scenarios are ignored in existing works. To address these two problems, we propose a benchmark, called BattleAgentBench, which defines seven sub-stages of three varying difficulty levels and conducts a fine-grained evaluation of language models in terms of single-agent scenario navigation capabilities, paired-agent task execution abilities, and multi-agent collaboration and competition capabilities. We conducted extensive evaluations on leading four closed-source and seven open-source models. Experimental results indicate that API-based models perform excellently on simple tasks but open-source small models struggle with simple tasks. Regarding difficult tasks that require collaborative and competitive abilities, although API-based models have demonstrated some collaborative capabilities, there is still enormous room for improvement.
Authors: Mustafa Omer Gul, Yoav Artzi
Abstract: Systems with both language comprehension and generation capabilities can benefit from the tight connection between the two. This work studies coupling comprehension and generation with focus on continually learning from interaction with users. We propose techniques to tightly integrate the two capabilities for both learning and inference. We situate our studies in two-player reference games, and deploy various models for thousands of interactions with human users, while learning from interaction feedback signals. We show dramatic improvements in performance over time, with comprehension-generation coupling leading to performance improvements up to 26% in absolute terms and up to 17% higher accuracies compared to a non-coupled system. Our analysis also shows coupling has substantial qualitative impact on the system's language, making it significantly more human-like.
Authors: Victor Dibia, Jingya Chen, Gagan Bansal, Suff Syed, Adam Fourney, Erkang Zhu, Chi Wang, Saleema Amershi
Abstract: Multi-agent systems, where multiple agents (generative AI models + tools) collaborate, are emerging as an effective pattern for solving long-running, complex tasks in numerous domains. However, specifying their parameters (such as models, tools, and orchestration mechanisms etc,.) and debugging them remains challenging for most developers. To address this challenge, we present AUTOGEN STUDIO, a no-code developer tool for rapidly prototyping, debugging, and evaluating multi-agent workflows built upon the AUTOGEN framework. AUTOGEN STUDIO offers a web interface and a Python API for representing LLM-enabled agents using a declarative (JSON-based) specification. It provides an intuitive drag-and-drop UI for agent workflow specification, interactive evaluation and debugging of workflows, and a gallery of reusable agent components. We highlight four design principles for no-code multi-agent developer tools and contribute an open-source implementation at https://github.com/microsoft/autogen/tree/main/samples/apps/autogen-studio
URLs: https://github.com/microsoft/autogen/tree/main/samples/apps/autogen-studio
Authors: Jinchuan Zhang, Tianqi Wan, Chong Mu, Guangxi Lu, Ling Tian
Abstract: Temporal Knowledge Graphs (TKGs) incorporate temporal information to reflect the dynamic structural knowledge and evolutionary patterns of real-world facts. Nevertheless, TKGs are still limited in downstream applications due to the problem of incompleteness. Consequently, TKG completion (also known as link prediction) has been widely studied, with recent research focusing on incorporating independent embeddings of time or combining them with entities and relations to form temporal representations. However, most existing methods overlook the impact of history from a multi-granularity aspect. The inherent semantics of human-defined temporal granularities, such as ordinal dates, reveal general patterns to which facts typically adhere. To counter this limitation, this paper proposes \textbf{L}earning \textbf{G}ranularity \textbf{Re}presentation (termed $\mathsf{LGRe}$) for TKG completion. It comprises two main components: Granularity Representation Learning (GRL) and Adaptive Granularity Balancing (AGB). Specifically, GRL employs time-specific multi-layer convolutional neural networks to capture interactions between entities and relations at different granularities. After that, AGB generates adaptive weights for these embeddings according to temporal semantics, resulting in expressive representations of predictions. Moreover, to reflect similar semantics of adjacent timestamps, a temporal loss function is introduced. Extensive experimental results on four event benchmarks demonstrate the effectiveness of $\mathsf{LGRe}$ in learning time-related representations. To ensure reproducibility, our code is available at https://github.com/KcAcoZhang/LGRe.
Authors: Xuanru Zhou, Anshul Kashyap, Steve Li, Ayati Sharma, Brittany Morin, David Baquirin, Jet Vonk, Zoe Ezzes, Zachary Miller, Maria Luisa Gorno Tempini, Jiachen Lian, Gopala Krishna Anumanchipalli
Abstract: Dysfluent speech detection is the bottleneck for disordered speech analysis and spoken language learning. Current state-of-the-art models are governed by rule-based systems which lack efficiency and robustness, and are sensitive to template design. In this paper, we propose YOLO-Stutter: a first end-to-end method that detects dysfluencies in a time-accurate manner. YOLO-Stutter takes imperfect speech-text alignment as input, followed by a spatial feature aggregator, and a temporal dependency extractor to perform region-wise boundary and class predictions. We also introduce two dysfluency corpus, VCTK-Stutter and VCTK-TTS, that simulate natural spoken dysfluencies including repetition, block, missing, replacement, and prolongation. Our end-to-end method achieves state-of-the-art performance with a minimum number of trainable parameters for on both simulated data and real aphasia speech. Code and datasets are open-sourced at https://github.com/rorizzz/YOLO-Stutter
Authors: Wenxuan Zhang, Philip H. S. Torr, Mohamed Elhoseiny, Adel Bibi
Abstract: Fine-tuning large language models (LLMs) on human preferences, typically through reinforcement learning from human feedback (RLHF), has proven successful in enhancing their capabilities. However, ensuring the safety of LLMs during the fine-tuning remains a critical concern, and mitigating the potential conflicts in safety and helpfulness is costly in RLHF. To address this issue, we propose a supervised learning framework called Bi-Factorial Preference Optimization (BFPO), which re-parameterizes a joint RLHF objective of both safety and helpfulness into a single supervised learning objective. In the supervised optimization, a labeling function is used to capture global preferences ranking to balance both safety and helpfulness. To evaluate BFPO, we develop a benchmark including comprehensive discriminative and generative tasks for helpfulness and harmlessness. The results indicate that our method significantly outperforms existing approaches in both safety and helpfulness. Moreover, BFPO eliminates the need for human prompting and annotation in LLM fine-tuning while achieving the same level of safety as methods that heavily rely on human labor, with less than 10% of the computational resources. The training recipes and models will be released.
Authors: Zhichao Wang, Bin Bi, Can Huang, Shiva Kumar Pentyala, Zixu James Zhu, Sitaram Asur, Na Claire Cheng
Abstract: An LLM is pretrained on trillions of tokens, but the pretrained LLM may still generate undesired responses. To solve this problem, alignment techniques such as RLHF, DPO and KTO are proposed. However, these alignment techniques have limitations. For example, RLHF requires training the reward model and policy separately, which is complex, time-consuming, memory intensive and unstable during training processes. DPO proposes a mapping between an optimal policy and a reward, greatly simplifying the training process of RLHF. However, it can not take full advantages of a reward model and it is limited to pairwise preference data. In this paper, we propose \textbf{UN}ified \textbf{A}lignment (UNA) which unifies RLHF/PPO, DPO and KTO. Firstly, we mathematically prove that given the classical RLHF objective, the optimal policy is induced by a generalize implicit reward function. With this novel mapping between a reward model and an optimal policy, UNA can 1. unify RLHF/PPO, DPO and KTO into a supervised learning of minimizing the difference between an implicit reward and an explicit reward; 2. outperform RLHF/PPO while simplify, stabilize, speed up and reduce memory burden of RL fine-tuning process; 3. accommodate different feedback types including pairwise, binary and scalar feedback. Downstream experiments show UNA outperforms DPO, KTO and RLHF.
Authors: Soumya Basu, Ankit Singh Rawat, Manzil Zaheer
Abstract: Modern ML systems increasingly augment input instances with additional relevant information to enhance final prediction. Despite growing interest in such retrieval-augmented models, their fundamental properties and training are not well understood. We propose a statistical framework to study such models with two components: 1) a {\em retriever} to identify the relevant information out of a large corpus via a data-dependent metric; and 2) a {\em predictor} that consumes the input instances along with the retrieved information to make the final predictions. We present a principled method for end-to-end training of both components and draw connections with various training approaches in the literature. Furthermore, we establish excess risk bounds for retrieval-augmented models while delineating the contributions of both retriever and predictor towards the model performance. We validate the utility of our proposed training methods along with the key takeaways from our statistical analysis on open domain question answering task where retrieval augmentation is important.
Authors: Yifan Liu, Yike Li, Dong Wang
Abstract: Media bias significantly shapes public perception by reinforcing stereotypes and exacerbating societal divisions. Prior research has often focused on isolated media bias dimensions such as \textit{political bias} or \textit{racial bias}, neglecting the complex interrelationships among various bias dimensions across different topic domains. Moreover, we observe that models trained on existing media bias benchmarks fail to generalize effectively on recent social media posts, particularly in certain bias identification tasks. This shortfall primarily arises because these benchmarks do not adequately reflect the rapidly evolving nature of social media content, which is characterized by shifting user behaviors and emerging trends. In response to these limitations, our research introduces a novel dataset collected from YouTube and Reddit over the past five years. Our dataset includes automated annotations for YouTube content across a broad spectrum of bias dimensions, such as gender, racial, and political biases, as well as hate speech, among others. It spans diverse domains including politics, sports, healthcare, education, and entertainment, reflecting the complex interplay of biases across different societal sectors. Through comprehensive statistical analysis, we identify significant differences in bias expression patterns and intra-domain bias correlations across these domains. By utilizing our understanding of the correlations among various bias dimensions, we lay the groundwork for creating advanced systems capable of detecting multiple biases simultaneously. Overall, our dataset advances the field of media bias identification, contributing to the development of tools that promote fairer media consumption. The comprehensive awareness of existing media bias fosters more ethical journalism, promotes cultural sensitivity, and supports a more informed and equitable public discourse.
Authors: Marc Canby, Adam Davies, Chirag Rastogi, Julia Hockenmaier
Abstract: Causal probing is an approach to interpreting foundation models, such as large language models, by training probes to recognize latent properties of interest from embeddings, intervening on probes to modify this representation, and analyzing the resulting changes in the model's behavior. While some recent works have cast doubt on the theoretical basis of several leading causal probing intervention methods, it has been unclear how to systematically and empirically evaluate their effectiveness in practice. To address this problem, we propose a general empirical analysis framework to evaluate the reliability of causal probing interventions, formally defining and quantifying two key causal probing desiderata: completeness (fully transforming the representation of the target property) and selectivity (minimally impacting other properties). Our formalism allows us to make the first direct comparisons between different families of causal probing methods (e.g., linear vs. nonlinear or counterfactual vs. nullifying interventions). We conduct extensive experiments across several leading methods, finding that (1) there is an inherent tradeoff between these criteria, and no method is able to consistently satisfy both at once; and (2) across the board, nullifying interventions are always far less complete than counterfactual interventions, indicating that nullifying methods may not be an effective approach to causal probing.
Authors: Zhihan Liu, Yubo Chai, Jianfeng Li
Abstract: The advent of Large Language Models (LLMs) has created new opportunities for the automation of scientific research, spanning both experimental processes and computational simulations. This study explores the feasibility of constructing an autonomous simulation agent (ASA) powered by LLM, through sophisticated API integration, to automate the entire research process, from experimental design, remote upload and simulation execution, data analysis, to report compilation. Using a simulation problem of polymer chain conformations as a case study, we assessed the performance of ASAs powered by different LLMs including GPT-4-Turbo. Our findings revealed that ASA-GPT-4o achieved near-flawless execution on designated research missions, underscoring the potential of LLMs to manage complete scientific investigations autonomously. The outlined automation can be iteratively performed up to twenty cycles without human intervention, illustrating the potential of LLMs for large-scale autonomous research endeavors. Additionally, we discussed the intrinsic traits of ASAs in managing extensive tasks, focusing on self-validation mechanisms and the balance between local attention and global oversight.
Authors: Sihang Li, Jian Huang, Jiaxi Zhuang, Yaorui Shi, Xiaochen Cai, Mingjun Xu, Xiang Wang, Linfeng Zhang, Guolin Ke, Hengxing Cai
Abstract: Scientific literature understanding is crucial for extracting targeted information and garnering insights, thereby significantly advancing scientific discovery. Despite the remarkable success of Large Language Models (LLMs), they face challenges in scientific literature understanding, primarily due to (1) a lack of scientific knowledge and (2) unfamiliarity with specialized scientific tasks. To develop an LLM specialized in scientific literature understanding, we propose a hybrid strategy that integrates continual pre-training (CPT) and supervised fine-tuning (SFT), to simultaneously infuse scientific domain knowledge and enhance instruction-following capabilities for domain-specific tasks.cIn this process, we identify two key challenges: (1) constructing high-quality CPT corpora, and (2) generating diverse SFT instructions. We address these challenges through a meticulous pipeline, including PDF text extraction, parsing content error correction, quality filtering, and synthetic instruction creation. Applying this strategy, we present a suite of LLMs: SciLitLLM, specialized in scientific literature understanding. These models demonstrate promising performance on scientific literature understanding benchmarks. Our contributions are threefold: (1) We present an effective framework that integrates CPT and SFT to adapt LLMs to scientific literature understanding, which can also be easily adapted to other domains. (2) We propose an LLM-based synthesis method to generate diverse and high-quality scientific instructions, resulting in a new instruction set -- SciLitIns -- for supervised fine-tuning in less-represented scientific domains. (3) SciLitLLM achieves promising performance improvements on scientific literature understanding benchmarks.
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 safety filter, designed based on the CBF, to the output generation of the baseline LLM, i.e., the sequence of the token, with the aim of intervening in the generated text. The overall text-generation system is implemented with Llama 3 and a RoBERTa model, and the source code is available at https://github.com/Mya-Mya/CBF-LLM. The experiment demonstrates its control ability and effectiveness in reducing the number of interventions needed for user-specified alignment tasks.
Authors: Lean Wang, Huazuo Gao, Chenggang Zhao, Xu Sun, Damai Dai
Abstract: For Mixture-of-Experts (MoE) models, an unbalanced expert load will lead to routing collapse or increased computational overhead. Existing methods commonly employ an auxiliary loss to encourage load balance, but a large auxiliary loss will introduce non-negligible interference gradients into training and thus impair the model performance. In order to control load balance while not producing undesired gradients during training, we propose Loss-Free Balancing, featured by an auxiliary-loss-free load balancing strategy. To be specific, before the top-K routing decision, Loss-Free Balancing will first apply an expert-wise bias to the routing scores of each expert. By dynamically updating the bias of each expert according to its recent load, Loss-Free Balancing can consistently maintain a balanced distribution of expert load. In addition, since Loss-Free Balancing does not produce any interference gradients, it also elevates the upper bound of model performance gained from MoE training. We validate the performance of Loss-Free Balancing on MoE models with up to 3B parameters trained on up to 200B tokens. Experimental results show that Loss-Free Balancing achieves both better performance and better load balance compared with traditional auxiliary-loss-controlled load balancing strategies.
Authors: Lefan Zhang, Xiaodan Wang, Yanhua Huang, Ruiwen Xu
Abstract: Speculative sampling has proven to be an effective solution to accelerate decoding from large language models, where the acceptance rate significantly determines the performance. Most previous works on improving the acceptance rate focus on aligned training and efficient decoding, implicitly paying less attention to the linkage of training and decoding. In this work, we first investigate the linkage of training and decoding for speculative sampling and then propose a solution named HArmonized Speculative Sampling (HASS). HASS improves the acceptance rate without extra inference overhead by harmonizing training and decoding on their objectives and contexts. Experiments on three LLaMA models demonstrate that HASS achieves 2.81x-3.65x wall-clock time speedup ratio averaging across three datasets, which is 8%-15% faster than EAGLE-2.
Authors: Jiaxing Huang, Jingyi Zhang
Abstract: Multimodal Large Language Models (MLLMs) mimic human perception and reasoning system by integrating powerful Large Language Models (LLMs) with various modality encoders (e.g., vision, audio), positioning LLMs as the "brain" and various modality encoders as sensory organs. This framework endows MLLMs with human-like capabilities, and suggests a potential pathway towards achieving artificial general intelligence (AGI). With the emergence of all-round MLLMs like GPT-4V and Gemini, a multitude of evaluation methods have been developed to assess their capabilities across different dimensions. This paper presents a systematic and comprehensive review of MLLM evaluation methods, covering the following key aspects: (1) the background of MLLMs and their evaluation; (2) "what to evaluate" that reviews and categorizes existing MLLM evaluation tasks based on the capabilities assessed, including general multimodal recognition, perception, reasoning and trustworthiness, and domain-specific applications such as socioeconomic, natural sciences and engineering, medical usage, AI agent, remote sensing, video and audio processing, 3D point cloud analysis, and others; (3) "where to evaluate" that summarizes MLLM evaluation benchmarks into general and specific benchmarks; (4) "how to evaluate" that reviews and illustrates MLLM evaluation steps and metrics; Our overarching goal is to provide valuable insights for researchers in the field of MLLM evaluation, thereby facilitating the development of more capable and reliable MLLMs. We emphasize that evaluation should be regarded as a critical discipline, essential for advancing the field of MLLMs.
Authors: Jiayi Gui, Yiming Liu, Jiale Cheng, Xiaotao Gu, Xiao Liu, Hongning Wang, Yuxiao Dong, Jie Tang, Minlie Huang
Abstract: Large Language Models (LLMs) have demonstrated notable capabilities across various tasks, showcasing complex problem-solving abilities. Understanding and executing complex rules, along with multi-step planning, are fundamental to logical reasoning and critical for practical LLM agents and decision-making systems. However, evaluating LLMs as effective rule-based executors and planners remains underexplored. In this paper, we introduce LogicGame, a novel benchmark designed to evaluate the comprehensive rule understanding, execution, and planning capabilities of LLMs. Unlike traditional benchmarks, LogicGame provides diverse games that contain a series of rules with an initial state, requiring models to comprehend and apply predefined regulations to solve problems. We create simulated scenarios in which models execute or plan operations to achieve specific outcomes. These game scenarios are specifically designed to distinguish logical reasoning from mere knowledge by relying exclusively on predefined rules. This separation allows for a pure assessment of rule-based reasoning capabilities. The evaluation considers not only final outcomes but also intermediate steps, providing a comprehensive assessment of model performance. Moreover, these intermediate steps are deterministic and can be automatically verified. LogicGame defines game scenarios with varying difficulty levels, from simple rule applications to complex reasoning chains, in order to offer a precise evaluation of model performance on rule understanding and multi-step execution. Utilizing LogicGame, we test various LLMs and identify notable shortcomings in their rule-based logical reasoning abilities.
Authors: Abu Adnan Sadi, Mohammad Ashrafuzzaman Khan, Lubaba Binte Saber
Abstract: As the field of artificial intelligence progresses, assistive technologies are becoming more widely used across all industries. The healthcare industry is no different, with numerous studies being done to develop assistive tools for healthcare professionals. Automatic diagnostic systems are one such beneficial tool that can assist with a variety of tasks, including collecting patient information, analyzing test results, and diagnosing patients. However, the idea of developing systems that can provide a differential diagnosis has been largely overlooked in most of these research studies. In this study, we propose a transformer-based approach for providing differential diagnoses based on a patient's age, sex, medical history, and symptoms. We use the DDXPlus dataset, which provides differential diagnosis information for patients based on 49 disease types. Firstly, we propose a method to process the tabular patient data from the dataset and engineer them into patient reports to make them suitable for our research. In addition, we introduce two data modification modules to diversify the training data and consequently improve the robustness of the models. We approach the task as a multi-label classification problem and conduct extensive experiments using four transformer models. All the models displayed promising results by achieving over 97% F1 score on the held-out test set. Moreover, we design additional behavioral tests to get a broader understanding of the models. In particular, for one of our test cases, we prepared a custom test set of 100 samples with the assistance of a doctor. The results on the custom set showed that our proposed data modification modules improved the model's generalization capabilities. We hope our findings will provide future researchers with valuable insights and inspire them to develop reliable systems for automatic differential diagnosis.
Authors: Uri Katz, Mosh Levy, Yoav Goldberg
Abstract: The exponential growth of scientific literature necessitates advanced tools for effective knowledge exploration. We present Knowledge Navigator, a system designed to enhance exploratory search abilities by organizing and structuring the retrieved documents from broad topical queries into a navigable, two-level hierarchy of named and descriptive scientific topics and subtopics. This structured organization provides an overall view of the research themes in a domain, while also enabling iterative search and deeper knowledge discovery within specific subtopics by allowing users to refine their focus and retrieve additional relevant documents. Knowledge Navigator combines LLM capabilities with cluster-based methods to enable an effective browsing method. We demonstrate our approach's effectiveness through automatic and manual evaluations on two novel benchmarks, CLUSTREC-COVID and SCITOC. Our code, prompts, and benchmarks are made publicly available.
Authors: Ganesh Prasath Ramani, Shirish Karande, Santhosh V, Yash Bhatia
Abstract: Large Language Models (LLMs) have emerged as formidable instruments capable of comprehending and producing human-like text. This paper explores the potential of LLMs, to shape human perspectives and subsequently influence their decisions on particular tasks. This capability finds applications in diverse domains such as Investment, Credit cards and Insurance, wherein they assist users in selecting appropriate insurance policies, investment plans, Credit cards, Retail, as well as in Behavioral Change Support Systems (BCSS). We present a sophisticated multi-agent framework wherein a consortium of agents operate in collaborative manner. The primary agent engages directly with users through persuasive dialogue, while the auxiliary agents perform tasks such as information retrieval, response analysis, development of persuasion strategies, and validation of facts. Empirical evidence from our experiments demonstrates that this collaborative methodology significantly enhances the persuasive efficacy of the LLM. We analyze user resistance to persuasive efforts continuously and counteract it by employing a combination of rule-based and LLM-based resistance-persuasion mapping techniques. We employ simulated personas and generate conversations in insurance, banking, and retail domains to evaluate the proficiency of large language models (LLMs) in recognizing, adjusting to, and influencing various personality types. Concurrently, we examine the resistance mechanisms employed by LLM simulated personas. Persuasion is quantified via measurable surveys before and after interaction, LLM-generated scores on conversation, and user decisions (purchase or non-purchase).
Authors: Yuncheng Yang, Yulei Qin, Tong Wu, Zihan Xu, Gang Li, Pengcheng Guo, Hang Shao, Yucheng Shi, Ke Li, Xing Sun, Jie Yang, Yun Gu
Abstract: The cultivation of expertise for large language models (LLMs) to solve tasks of specific areas often requires special-purpose tuning with calibrated behaviors on the expected stable outputs. To avoid huge cost brought by manual preparation of instruction datasets and training resources up to hundreds of hours, the exploitation of open knowledge including a wealth of low rank adaptation (LoRA) models and instruction datasets serves as a good starting point. However, existing methods on model and data selection focus on the performance of general-purpose capabilities while neglecting the knowledge gap exposed in domain-specific deployment. In the present study, we propose to bridge such gap by introducing few human-annotated samples (i.e., K-shot) for advancing task expertise of LLMs with open knowledge. Specifically, we develop an efficient and scalable pipeline to cost-efficiently produce task experts where K-shot data intervene in selecting the most promising expert candidates and the task-relevant instructions. A mixture-of-expert (MoE) system is built to make the best use of individual-yet-complementary knowledge between multiple experts. We unveil the two keys to the success of a MoE system, 1) the abidance by K-shot, and 2) the insistence on diversity. For the former, we ensure that models that truly possess problem-solving abilities on K-shot are selected rather than those blind guessers. Besides, during data selection, instructions that share task-relevant contexts with K-shot are prioritized. For the latter, we highlight the diversity of constituting experts and that of the fine-tuning instructions throughout the model and data selection process. Extensive experimental results confirm the superiority of our approach over existing methods on utilization of open knowledge across various tasks. Codes and models will be released later.
Authors: Yuan Tang, Xu Han, Xianzhi Li, Qiao Yu, Jinfeng Xu, Yixue Hao, Long Hu, Min Chen
Abstract: Enabling Large Language Models (LLMs) to comprehend the 3D physical world remains a significant challenge. Due to the lack of large-scale 3D-text pair datasets, the success of LLMs has yet to be replicated in 3D understanding. In this paper, we rethink this issue and propose a new task: 3D Data-Efficient Point-Language Understanding. The goal is to enable LLMs to achieve robust 3D object understanding with minimal 3D point cloud and text data pairs. To address this task, we introduce GreenPLM, which leverages more text data to compensate for the lack of 3D data. First, inspired by using CLIP to align images and text, we utilize a pre-trained point cloud-text encoder to map the 3D point cloud space to the text space. This mapping leaves us to seamlessly connect the text space with LLMs. Once the point-text-LLM connection is established, we further enhance text-LLM alignment by expanding the intermediate text space, thereby reducing the reliance on 3D point cloud data. Specifically, we generate 6M free-text descriptions of 3D objects, and design a three-stage training strategy to help LLMs better explore the intrinsic connections between different modalities. To achieve efficient modality alignment, we design a zero-parameter cross-attention module for token pooling. Extensive experimental results show that GreenPLM requires only 12% of the 3D training data used by existing state-of-the-art models to achieve superior 3D understanding. Remarkably, GreenPLM also achieves competitive performance using text-only data. The code and weights are available at: https://github.com/TangYuan96/GreenPLM.
Authors: Zhenpeng Su, Xing Wu, Wei Zhou, Guangyuan Ma, Songlin Hu
Abstract: ChatGPT has garnered significant interest due to its impressive performance; however, there is growing concern about its potential risks, particularly in the detection of AI-generated content (AIGC), which is often challenging for untrained individuals to identify. Current datasets used for detecting ChatGPT-generated text primarily focus on question-answering tasks, often overlooking tasks with semantic-invariant properties, such as summarization, translation, and paraphrasing. In this paper, we demonstrate that detecting model-generated text in semantic-invariant tasks is more challenging. To address this gap, we introduce a more extensive and comprehensive dataset that incorporates a wider range of tasks than previous work, including those with semantic-invariant properties.
Authors: Andreas Madsen, Siva Reddy, Sarath Chandar
Abstract: A common approach to explaining NLP models is to use importance measures that express which tokens are important for a prediction. Unfortunately, such explanations are often wrong despite being persuasive. Therefore, it is essential to measure their faithfulness. One such metric is if tokens are truly important, then masking them should result in worse model performance. However, token masking introduces out-of-distribution issues, and existing solutions that address this are computationally expensive and employ proxy models. Furthermore, other metrics are very limited in scope. This work proposes an inherently faithfulness measurable model that addresses these challenges. This is achieved using a novel fine-tuning method that incorporates masking, such that masking tokens become in-distribution by design. This differs from existing approaches, which are completely model-agnostic but are inapplicable in practice. We demonstrate the generality of our approach by applying it to 16 different datasets and validate it using statistical in-distribution tests. The faithfulness is then measured with 9 different importance measures. Because masking is in-distribution, importance measures that themselves use masking become consistently more faithful. Additionally, because the model makes faithfulness cheap to measure, we can optimize explanations towards maximal faithfulness; thus, our model becomes indirectly inherently explainable.
Authors: Lorenzo Lupo, Oscar Magnusson, Dirk Hovy, Elin Naurin, Lena W\"angnerud
Abstract: Recent advances in large language models (LLMs) like GPT-3.5 and GPT-4 promise automation with better results and less programming, opening up new opportunities for text analysis in political science. In this study, we evaluate LLMs on three original coding tasks involving typical complexities encountered in political science settings: a non-English language, legal and political jargon, and complex labels based on abstract constructs. Along the paper, we propose a practical workflow to optimize the choice of the model and the prompt. We find that the best prompting strategy consists of providing the LLMs with a detailed codebook, as the one provided to human coders. In this setting, an LLM can be as good as or possibly better than a human annotator while being much faster, considerably cheaper, and much easier to scale to large amounts of text. We also provide a comparison of GPT and popular open-source LLMs, discussing the trade-offs in the model's choice. Our software allows LLMs to be easily used as annotators and is publicly available: https://github.com/lorelupo/pappa.
Authors: Yang Zhao, Li Du, Xiao Ding, Kai Xiong, Zhouhao Sun, Jun Shi, Ting Liu, Bing Qin
Abstract: Through pretraining on a corpus with various sources, Large Language Models (LLMs) have gained impressive performance. However, the impact of each component of the pretraining corpus remains opaque. As a result, the organization of the pretraining corpus is still empirical and may deviate from the optimal. To address this issue, we systematically analyze the impact of 48 datasets from 5 major categories of pretraining data of LLMs and measure their impacts on LLMs using benchmarks about nine major categories of model capabilities. Our analyses provide empirical results about the contribution of multiple corpora on the performances of LLMs, along with their joint impact patterns, including complementary, orthogonal, and correlational relationships. We also identify a set of ``high-impact data'' such as Books that is significantly related to a set of model capabilities. These findings provide insights into the organization of data to support more efficient pretraining of LLMs.
Authors: Grgur Kova\v{c}, R\'emy Portelas, Masataka Sawayama, Peter Ford Dominey, Pierre-Yves Oudeyer
Abstract: The standard way to study Large Language Models (LLMs) with benchmarks or psychology questionnaires is to provide many different queries from similar minimal contexts (e.g. multiple choice questions). However, due to LLMs' highly context-dependent nature, conclusions from such minimal-context evaluations may be little informative about the model's behavior in deployment (where it will be exposed to many new contexts). We argue that context-dependence (specifically, value stability) should be studied as a specific property of LLMs and used as another dimension of LLM comparison (alongside others such as cognitive abilities, knowledge, or model size). We present a case-study on the stability of value expression over different contexts (simulated conversations on different topics) as measured using a standard psychology questionnaire (PVQ) and on behavioral downstream tasks. Reusing methods from psychology, we study Rank-order stability on the population (interpersonal) level, and Ipsative stability on the individual (intrapersonal) level. We consider two settings (with and without instructing LLMs to simulate particular personas), two simulated populations, and three downstream tasks. We observe consistent trends in the stability of models and model families - Mixtral, Mistral, GPT-3.5 and Qwen families are more stable than LLaMa-2 and Phi. The consistency of these trends implies that some models exhibit higher value stability than others, and that stability can be estimated with the set of introduced methodological tools. When instructed to simulate particular personas, LLMs exhibit low Rank-order stability, which further diminishes with conversation length. This highlights the need for future research on LLMs that coherently simulate different personas. This paper provides a foundational step in that direction, and, to our knowledge, it is the first study of value stability in LLMs.
Authors: Anchun Gui, Jian Li, Yong Dai, Nan Du, Han Xiao
Abstract: Tool-augmented large language models (LLMs) are attracting widespread attention when accessing up-to-date knowledge and alleviating hallucination issues. Nowadays, advanced closed-source LLMs (e.g., ChatGPT) have demonstrated surprising tool-usage capabilities through prompting and in-context learning techniques. To empower the capabilities of open-source LLMs (e.g., LLaMA) in manipulating tools, current efforts focus on either template-driven or token-triggered tool-usage. However, the former hampers LLMs' flexibility to address diverse user's queries due to constrained tool interactions, while the latter limits the generalizability when engaging with new tools, since tool-usage learning is based on task- and tool-specific datasets. To alleviate these concerns, in this paper, we propose a decision-aware and generalizable tool-usage framework (DEER). Specifically, we first construct the tool-usage samples with multiple decision branches via an automatic generation pipeline, thereby inspiring the decision-making awareness of LLMs under diverse scenarios. Meanwhile, we propose a novel tool sampling strategy to enhance the generalizability of LLMs over unseen tools. Extensive experiments demonstrate that our proposed DEER is effective and significantly outperforms baselines across various datasets.
Authors: Amir Zeldes, Tatsuya Aoyama, Yang Janet Liu, Siyao Peng, Debopam Das, Luke Gessler
Abstract: In this article we present Enhanced Rhetorical Structure Theory (eRST), a new theoretical framework for computational discourse analysis, based on an expansion of Rhetorical Structure Theory (RST). The framework encompasses discourse relation graphs with tree-breaking, non-projective and concurrent relations, as well as implicit and explicit signals which give explainable rationales to our analyses. We survey shortcomings of RST and other existing frameworks, such as Segmented Discourse Representation Theory (SDRT), the Penn Discourse Treebank (PDTB) and Discourse Dependencies, and address these using constructs in the proposed theory. We provide annotation, search and visualization tools for data, and present and evaluate a freely available corpus of English annotated according to our framework, encompassing 12 spoken and written genres with over 200K tokens. Finally, we discuss automatic parsing, evaluation metrics and applications for data in our framework.
Authors: R. Michael Alvarez, Jacob Morrier
Abstract: This paper introduces a new approach for measuring the quality of answers in political question-and-answer sessions. We propose to measure answer quality based on the degree to which it allows to infer the initial question accurately. This measure of answer quality reflects how well the answer engages with and addresses the initial question. Drawing an analogy with semantic search, we demonstrate that this measurement approach can be implemented by fine-tuning a large language model on the corpus of observed questions and answers without additional labeled data. We showcase our approach within the context of the Question Period in the Canadian House of Commons, providing valuable insights into the correlates of answer quality. Our findings reveal significant variations in answer quality based on the party affiliation of the members of Parliament asking the question. Additionally, we find a meaningful correlation between answer quality and the topic raised in the question.
Authors: David M. Markowitz
Abstract: This paper evaluated the effectiveness of using generative AI to simplify science communication and enhance the public's understanding of science. By comparing lay summaries of journal articles from PNAS, yoked to those generated by AI, this work first assessed linguistic simplicity differences across such summaries and public perceptions in follow-up experiments. Specifically, Study 1a analyzed simplicity features of PNAS abstracts (scientific summaries) and significance statements (lay summaries), observing that lay summaries were indeed linguistically simpler, but effect size differences were small. Study 1b used a large language model, GPT-4, to create significance statements based on paper abstracts and this more than doubled the average effect size without fine-tuning. Study 2 experimentally demonstrated that simply-written GPT summaries facilitated more favorable perceptions of scientists (they were perceived as more credible and trustworthy, but less intelligent) than more complexly-written human PNAS summaries. Crucially, Study 3 experimentally demonstrated that participants comprehended scientific writing better after reading simple GPT summaries compared to complex PNAS summaries. In their own words, participants also summarized scientific papers in a more detailed and concrete manner after reading GPT summaries compared to PNAS summaries of the same article. AI has the potential to engage scientific communities and the public via a simple language heuristic, advocating for its integration into scientific dissemination for a more informed society.
Authors: Guang Lin, Qibin Zhao
Abstract: Over the past two years, the use of large language models (LLMs) has advanced rapidly. While these LLMs offer considerable convenience, they also raise security concerns, as LLMs are vulnerable to adversarial attacks by some well-designed textual perturbations. In this paper, we introduce a novel defense technique named Large LAnguage MOdel Sentinel (LLAMOS), which is designed to enhance the adversarial robustness of LLMs by purifying the adversarial textual examples before feeding them into the target LLM. Our method comprises two main components: a) Agent instruction, which can simulate a new agent for adversarial defense, altering minimal characters to maintain the original meaning of the sentence while defending against attacks; b) Defense guidance, which provides strategies for modifying clean or adversarial examples to ensure effective defense and accurate outputs from the target LLMs. Remarkably, the defense agent demonstrates robust defensive capabilities even without learning from adversarial examples. Additionally, we conduct an intriguing adversarial experiment where we develop two agents, one for defense and one for attack, and engage them in mutual confrontation. During the adversarial interactions, neither agent completely beat the other. Extensive experiments on both open-source and closed-source LLMs demonstrate that our method effectively defends against adversarial attacks, thereby enhancing adversarial robustness.
Authors: Ruisi Cai, Saurav Muralidharan, Greg Heinrich, Hongxu Yin, Zhangyang Wang, Jan Kautz, Pavlo Molchanov
Abstract: Training modern LLMs is extremely resource intensive, and customizing them for various deployment scenarios characterized by limited compute and memory resources through repeated training is impractical. In this paper, we introduce Flextron, a network architecture and post-training model optimization framework supporting flexible model deployment. The Flextron architecture utilizes a nested elastic structure to rapidly adapt to specific user-defined latency and accuracy targets during inference with no additional fine-tuning required. It is also input-adaptive, and can automatically route tokens through its sub-networks for improved performance and efficiency. We present a sample-efficient training method and associated routing algorithms for systematically transforming an existing trained LLM into a Flextron model. We evaluate Flextron on the GPT-3 and LLama-2 family of LLMs, and demonstrate superior performance over multiple end-to-end trained variants and other state-of-the-art elastic networks, all with a single pretraining run that consumes a mere 7.63% tokens compared to original pretraining.
Authors: Aldan Creo, Shushanta Pudasaini
Abstract: The advent of large language models (LLMs) has enabled the generation of text that increasingly exhibits human-like characteristics. As the detection of such content is of significant importance, numerous studies have been conducted with the aim of developing reliable AI-generated text detectors. These detectors have demonstrated promising results on test data, but recent research has revealed that they can be circumvented by employing different techniques. In this paper, we present homoglyph-based attacks ($a \rightarrow {\alpha}$) as a means of circumventing existing detectors. A comprehensive evaluation was conducted to assess the effectiveness of these attacks on seven detectors, including ArguGPT, Binoculars, DetectGPT, Fast-DetectGPT, Ghostbuster, OpenAI's detector, and watermarking techniques, on five different datasets. Our findings demonstrate that homoglyph-based attacks can effectively circumvent state-of-the-art detectors, leading them to classify all texts as either AI-generated or human-written (decreasing the average Matthews Correlation Coefficient from 0.64 to -0.01). We then examine the effectiveness of these attacks by analyzing how homoglyphs impact different families of detectors. Finally, we discuss the implications of these findings and potential defenses against such attacks.
Authors: Jingbo Shang, Zai Zheng, Jiale Wei, Xiang Ying, Felix Tao, Mindverse Team
Abstract: Large language models (LLMs) have demonstrated the world with the sparks of artificial general intelligence (AGI). One opinion, especially from some startups working on LLMs, argues that an LLM with nearly unlimited context length can realize AGI. However, they might be too optimistic about the long-context capability of (existing) LLMs -- (1) Recent literature has shown that their effective context length is significantly smaller than their claimed context length; and (2) Our reasoning-in-a-haystack experiments further demonstrate that simultaneously finding the relevant information from a long context and conducting (simple) reasoning is nearly impossible. In this paper, we envision a pathway from LLMs to AGI through the integration of \emph{memory}. We believe that AGI should be a system where LLMs serve as core processors. In addition to raw data, the memory in this system would store a large number of important conclusions derived from reasoning processes. Compared with retrieval-augmented generation (RAG) that merely processing raw data, this approach not only connects semantically related information closer, but also simplifies complex inferences at the time of querying. As an intermediate stage, the memory will likely be in the form of natural language descriptions, which can be directly consumed by users too. Ultimately, every agent/person should have its own large personal model, a deep neural network model (thus \emph{AI-native}) that parameterizes and compresses all types of memory, even the ones cannot be described by natural languages. Finally, we discuss the significant potential of AI-native memory as the transformative infrastructure for (proactive) engagement, personalization, distribution, and social in the AGI era, as well as the incurred privacy and security challenges with preliminary solutions.
Authors: Weiming Li, Manni Duan, Dong An, Yan Shao
Abstract: Large language models (LLMs) demonstrate extraordinary abilities in a wide range of natural language processing (NLP) tasks. In this paper, we show that, beyond text understanding capability, LLMs are capable of processing text layouts that are denoted by spatial markers. They are able to answer questions that require explicit spatial perceiving and reasoning, while a drastic performance drop is observed when the spatial markers from the original data are excluded. We perform a series of experiments with the GPT-3.5, Baichuan2, Llama2 and ChatGLM3 models on various types of layout-sensitive datasets for further analysis. The experimental results reveal that the layout understanding ability of LLMs is mainly introduced by the coding data for pretraining, which is further enhanced at the instruction-tuning stage. In addition, layout understanding can be enhanced by integrating low-cost, auto-generated data approached by a novel text game. Finally, we show that layout understanding ability is beneficial for building efficient visual question-answering (VQA) systems.
Authors: Xihui Lin, Yunan Zhang, Suyu Ge, Barun Patra, Vishrav Chaudhary, Hao Peng, Xia Song
Abstract: Existing LLM training and inference frameworks struggle in boosting efficiency with sparsity while maintaining the integrity of context and model architecture. Inspired by the sharding concept in database and the fact that attention parallelizes over heads on accelerators, we propose Sparsely-Sharded (S2) Attention, an attention algorithm that allocates heterogeneous context partitions for different attention heads to divide and conquer. S2-Attention enforces each attention head to only attend to a partition of contexts following a strided sparsity pattern, while the full context is preserved as the union of all the shards. As attention heads are processed in separate thread blocks, the context reduction for each head can thus produce end-to-end speed-up and memory reduction. At inference, LLMs trained with S2-Attention can then take the KV cache reduction as free meals with guaranteed model quality preserve. In experiments, we show S2-Attentioncan provide as much as (1) 25.3X wall-clock attention speed-up over FlashAttention-2, resulting in 6X reduction in end-to-end training time and 10X inference latency, (2) on-par model training quality compared to default attention, (3)perfect needle retrieval accuracy over 32K context window. On top of the algorithm, we build DKernel, an LLM training and inference kernel library that allows users to customize sparsity patterns for their own models. We open-sourced DKerneland make it compatible with Megatron, Pytorch, and vLLM.
Authors: Arkadeep Baksi, Rahul Singh, Tarun Joshi
Abstract: The advent of transformer-based architectures and large language models (LLMs) have significantly advanced the performance of natural language processing (NLP) models. Since these LLMs are trained on huge corpuses of data from the web and other sources, there has been a major concern about harmful prejudices that may potentially be transferred from the data. In many applications, these pre-trained LLMs are fine-tuned on task specific datasets, which can further contribute to biases. This paper studies the extent of biases absorbed by LLMs during pre-training as well as task-specific behaviour after fine-tuning. We found that controlled interventions on pre-trained LLMs, prior to fine-tuning, have minimal effect on lowering biases in classifiers. However, the biases present in domain-specific datasets play a much bigger role, and hence mitigating them at this stage has a bigger impact. While pre-training does matter, but after the model has been pre-trained, even slight changes to co-occurrence rates in the fine-tuning dataset has a significant effect on the bias of the model.
Authors: Robert Wolfe, Alexis Hiniker, Bill Howe
Abstract: This research introduces the Multilevel Embedding Association Test (ML-EAT), a method designed for interpretable and transparent measurement of intrinsic bias in language technologies. The ML-EAT addresses issues of ambiguity and difficulty in interpreting the traditional EAT measurement by quantifying bias at three levels of increasing granularity: the differential association between two target concepts with two attribute concepts; the individual effect size of each target concept with two attribute concepts; and the association between each individual target concept and each individual attribute concept. Using the ML-EAT, this research defines a taxonomy of EAT patterns describing the nine possible outcomes of an embedding association test, each of which is associated with a unique EAT-Map, a novel four-quadrant visualization for interpreting the ML-EAT. Empirical analysis of static and diachronic word embeddings, GPT-2 language models, and a CLIP language-and-image model shows that EAT patterns add otherwise unobservable information about the component biases that make up an EAT; reveal the effects of prompting in zero-shot models; and can also identify situations when cosine similarity is an ineffective metric, rendering an EAT unreliable. Our work contributes a method for rendering bias more observable and interpretable, improving the transparency of computational investigations into human minds and societies.
Authors: Konstantinos Kogkalidis, Stergios Chatzikyriakidis
Abstract: This paper is a critical reflection on the epistemic culture of contemporary computational linguistics, framed in the context of its growing obsession with tables with numbers. We argue against tables with numbers on the basis of their epistemic irrelevance, their environmental impact, their role in enabling and exacerbating social inequalities, and their deep ties to commercial applications and profit-driven research. We substantiate our arguments with empirical evidence drawn from a meta-analysis of computational linguistics research over the last decade.
Authors: Weijian Xie, Xuefeng Liang, Yuhui Liu, Kaihua Ni, Hong Cheng, Zetian Hu
Abstract: Large Language Models (LLMs) have greatly contributed to the development of adaptive intelligent agents and are positioned as an important way to achieve Artificial General Intelligence (AGI). However, LLMs are prone to produce factually incorrect information and often produce "phantom" content that undermines their reliability, which poses a serious challenge for their deployment in real-world scenarios. Enhancing LLMs by combining external databases and information retrieval mechanisms is an effective path. To address the above challenges, we propose a new approach called WeKnow-RAG, which integrates Web search and Knowledge Graphs into a "Retrieval-Augmented Generation (RAG)" system. First, the accuracy and reliability of LLM responses are improved by combining the structured representation of Knowledge Graphs with the flexibility of dense vector retrieval. WeKnow-RAG then utilizes domain-specific knowledge graphs to satisfy a variety of queries and domains, thereby improving performance on factual information and complex reasoning tasks by employing multi-stage web page retrieval techniques using both sparse and dense retrieval methods. Our approach effectively balances the efficiency and accuracy of information retrieval, thus improving the overall retrieval process. Finally, we also integrate a self-assessment mechanism for the LLM to evaluate the trustworthiness of the answers it generates. Our approach proves its outstanding effectiveness in a wide range of offline experiments and online submissions.
Authors: Jing Tang, Quanlu Jia, Yuqiang Xie, Zeyu Gong, Xiang Wen, Jiayi Zhang, Yalong Guo, Guibin Chen, Jiangping Yang
Abstract: Generating high-quality shooting scripts containing information such as scene and shot language is essential for short drama script generation. We collect 6,660 popular short drama episodes from the Internet, each with an average of 100 short episodes, and the total number of short episodes is about 80,000, with a total duration of about 2,000 hours and totaling 10 terabytes (TB). We perform keyframe extraction and annotation on each episode to obtain about 10,000,000 shooting scripts. We perform 100 script restorations on the extracted shooting scripts based on our self-developed large short drama generation model SkyReels. This leads to a dataset containing 1,000,000,000 pairs of scripts and shooting scripts for short dramas, called SkyScript-100M. We compare SkyScript-100M with the existing dataset in detail and demonstrate some deeper insights that can be achieved based on SkyScript-100M. Based on SkyScript-100M, researchers can achieve several deeper and more far-reaching script optimization goals, which may drive a paradigm shift in the entire field of text-to-video and significantly advance the field of short drama video generation. The data and code are available at https://github.com/vaew/SkyScript-100M.
Authors: Simon Kurz, Jian-Jia Chen, Lucie Flek, Zhixue Zhao
Abstract: Recent advances in large language model (LLM) pruning have shown state-of-the-art compression results in post-training and retraining-free settings while maintaining high predictive performance. However, such research mainly considers calibrating pruning using English text, despite the multilingual nature of modern LLMs and their frequent uses in non-English languages. In this paper, we set out to explore effective strategies for calibrating the pruning of multilingual language models. We present the first comprehensive empirical study, comparing different calibration languages for pruning multilingual models across diverse tasks, models, and state-of-the-art pruning techniques. Our results present practical suggestions, for example, calibrating in the target language can efficiently yield lower perplexity, but does not necessarily benefit downstream tasks. Our further analysis experiments unveil that calibration in the target language mainly contributes to preserving language-specific features related to fluency and coherence, but might not contribute to capturing language-agnostic features such as language understanding and reasoning. Last, we provide practical recommendations for future practitioners.
Authors: Liuchang Xu, Shuo Zhao, Qingming Lin, Luyao Chen, Qianqian Luo, Sensen Wu, Xinyue Ye, Hailin Feng, Zhenhong Du
Abstract: The advent of large language models such as ChatGPT, Gemini, and others has underscored the importance of evaluating their diverse capabilities, ranging from natural language understanding to code generation. However, their performance on spatial tasks has not been comprehensively assessed. This study addresses this gap by introducing a novel multi-task spatial evaluation dataset, designed to systematically explore and compare the performance of several advanced models on spatial tasks. The dataset encompasses twelve distinct task types, including spatial understanding and path planning, each with verified, accurate answers. We evaluated multiple models, including OpenAI's gpt-3.5-turbo, gpt-4o, and ZhipuAI's glm-4, through a two-phase testing approach. Initially, we conducted zero-shot testing, followed by categorizing the dataset by difficulty and performing prompt tuning tests. Results indicate that gpt-4o achieved the highest overall accuracy in the first phase, with an average of 71.3%. Although moonshot-v1-8k slightly underperformed overall, it surpassed gpt-4o in place name recognition tasks. The study also highlights the impact of prompt strategies on model performance in specific tasks. For example, the Chain-of-Thought (COT) strategy increased gpt-4o's accuracy in path planning from 12.4% to 87.5%, while a one-shot strategy enhanced moonshot-v1-8k's accuracy in mapping tasks from 10.1% to 76.3%.
Authors: Wazir Ali, Sampo Pyysalo
Abstract: Large Language Models (LLMs) have gained significant attention due to their high performance on a wide range of natural language tasks since the release of ChatGPT. The LLMs learn to understand and generate language by training billions of model parameters on vast volumes of text data. Despite being a relatively new field, LLM research is rapidly advancing in various directions. In this paper, we present an overview of LLM families, including LLaMA, PaLM, GPT, and MoE, and the methods developed to create and enhance LLMs for official European Union (EU) languages. We provide a comprehensive summary of common monolingual and multilingual datasets used for pretraining large language models.
Authors: Yixuan Qiao, Hao Chen, Jun Wang, Tuozhen Liu, Xianbin Ye, Xin Tang, Rui Fang, Peng Gao, Wenfeng Xie, Guotong Xie
Abstract: This paper describes the PASH participation in TREC 2021 Deep Learning Track. In the recall stage, we adopt a scheme combining sparse and dense retrieval method. In the multi-stage ranking phase, point-wise and pair-wise ranking strategies are used one after another based on model continual pre-trained on general knowledge and document-level data. Compared to TREC 2020 Deep Learning Track, we have additionally introduced the generative model T5 to further enhance the performance.
Authors: Xiang Li, Feng Ruan, Huiyuan Wang, Qi Long, Weijie J. Su
Abstract: Since ChatGPT was introduced in November 2022, embedding (nearly) unnoticeable statistical signals into text generated by large language models (LLMs), also known as watermarking, has been used as a principled approach to provable detection of LLM-generated text from its human-written counterpart. In this paper, we introduce a general and flexible framework for reasoning about the statistical efficiency of watermarks and designing powerful detection rules. Inspired by the hypothesis testing formulation of watermark detection, our framework starts by selecting a pivotal statistic of the text and a secret key -- provided by the LLM to the verifier -- to enable controlling the false positive rate (the error of mistakenly detecting human-written text as LLM-generated). Next, this framework allows one to evaluate the power of watermark detection rules by obtaining a closed-form expression of the asymptotic false negative rate (the error of incorrectly classifying LLM-generated text as human-written). Our framework further reduces the problem of determining the optimal detection rule to solving a minimax optimization program. We apply this framework to two representative watermarks -- one of which has been internally implemented at OpenAI -- and obtain several findings that can be instrumental in guiding the practice of implementing watermarks. In particular, we derive optimal detection rules for these watermarks under our framework. These theoretically derived detection rules are demonstrated to be competitive and sometimes enjoy a higher power than existing detection approaches through numerical experiments.
Authors: Aleksandar Botev, Soham De, Samuel L Smith, Anushan Fernando, George-Cristian Muraru, Ruba Haroun, Leonard Berrada, Razvan Pascanu, Pier Giuseppe Sessa, Robert Dadashi, L\'eonard Hussenot, Johan Ferret, Sertan Girgin, Olivier Bachem, Alek Andreev, Kathleen Kenealy, Thomas Mesnard, Cassidy Hardin, Surya Bhupatiraju, Shreya Pathak, Laurent Sifre, Morgane Rivi\`ere, Mihir Sanjay Kale, Juliette Love, Pouya Tafti, Armand Joulin, Noah Fiedel, Evan Senter, Yutian Chen, Srivatsan Srinivasan, Guillaume Desjardins, David Budden, Arnaud Doucet, Sharad Vikram, Adam Paszke, Trevor Gale, Sebastian Borgeaud, Charlie Chen, Andy Brock, Antonia Paterson, Jenny Brennan, Meg Risdal, Raj Gundluru, Nesh Devanathan, Paul Mooney, Nilay Chauhan, Phil Culliton, Luiz Gustavo Martins, Elisa Bandy, David Huntsperger, Glenn Cameron, Arthur Zucker, Tris Warkentin, Ludovic Peran, Minh Giang, Zoubin Ghahramani, Cl\'ement Farabet, Koray Kavukcuoglu, Demis Hassabis, Raia Hadsell, Yee Whye Teh, Nando de Frietas
Abstract: We introduce RecurrentGemma, a family of open language models which uses Google's novel Griffin architecture. Griffin combines linear recurrences with local attention to achieve excellent performance on language. It has a fixed-sized state, which reduces memory use and enables efficient inference on long sequences. We provide two sizes of models, containing 2B and 9B parameters, and provide pre-trained and instruction tuned variants for both. Our models achieve comparable performance to similarly-sized Gemma baselines despite being trained on fewer tokens.
Authors: Robert Wolfe, Aayushi Dangol, Alexis Hiniker, Bill Howe
Abstract: Multimodal AI models capable of associating images and text hold promise for numerous domains, ranging from automated image captioning to accessibility applications for blind and low-vision users. However, uncertainty about bias has in some cases limited their adoption and availability. In the present work, we study 43 CLIP vision-language models to determine whether they learn human-like facial impression biases, and we find evidence that such biases are reflected across three distinct CLIP model families. We show for the first time that the the degree to which a bias is shared across a society predicts the degree to which it is reflected in a CLIP model. Human-like impressions of visually unobservable attributes, like trustworthiness and sexuality, emerge only in models trained on the largest dataset, indicating that a better fit to uncurated cultural data results in the reproduction of increasingly subtle social biases. Moreover, we use a hierarchical clustering approach to show that dataset size predicts the extent to which the underlying structure of facial impression bias resembles that of facial impression bias in humans. Finally, we show that Stable Diffusion models employing CLIP as a text encoder learn facial impression biases, and that these biases intersect with racial biases in Stable Diffusion XL-Turbo. While pretrained CLIP models may prove useful for scientific studies of bias, they will also require significant dataset curation when intended for use as general-purpose models in a zero-shot setting.
Authors: Le Xue, Manli Shu, Anas Awadalla, Jun Wang, An Yan, Senthil Purushwalkam, Honglu Zhou, Viraj Prabhu, Yutong Dai, Michael S Ryoo, Shrikant Kendre, Jieyu Zhang, Can Qin, Shu Zhang, Chia-Chih Chen, Ning Yu, Juntao Tan, Tulika Manoj Awalgaonkar, Shelby Heinecke, Huan Wang, Yejin Choi, Ludwig Schmidt, Zeyuan Chen, Silvio Savarese, Juan Carlos Niebles, Caiming Xiong, Ran Xu
Abstract: This report introduces xGen-MM (also known as BLIP-3), a framework for developing Large Multimodal Models (LMMs). The framework comprises meticulously curated datasets, a training recipe, model architectures, and a resulting suite of LMMs. xGen-MM, short for xGen-MultiModal, expands the Salesforce xGen initiative on foundation AI models. Our models undergo rigorous evaluation across a range of tasks, including both single and multi-image benchmarks. Our pre-trained base model exhibits strong in-context learning capabilities and the instruction-tuned model demonstrates competitive performance among open-source LMMs with similar model sizes. In addition, we introduce a safety-tuned model with DPO, aiming to mitigate harmful behaviors such as hallucinations and improve safety. We open-source our models, curated large-scale datasets, and our fine-tuning codebase to facilitate further advancements in LMM research. Associated resources will be available on our project page above.
Authors: Dongchao Yang, Rongjie Huang, Yuanyuan Wang, Haohan Guo, Dading Chong, Songxiang Liu, Xixin Wu, Helen Meng
Abstract: Scaling Text-to-speech (TTS) to large-scale datasets has been demonstrated as an effective method for improving the diversity and naturalness of synthesized speech. At the high level, previous large-scale TTS models can be categorized into either Auto-regressive (AR) based (\textit{e.g.}, VALL-E) or Non-auto-regressive (NAR) based models (\textit{e.g.}, NaturalSpeech 2/3). Although these works demonstrate good performance, they still have potential weaknesses. For instance, AR-based models are plagued by unstable generation quality and slow generation speed; meanwhile, some NAR-based models need phoneme-level duration alignment information, thereby increasing the complexity of data pre-processing, model design, and loss design. In this work, we build upon our previous publication by implementing a simple and efficient non-autoregressive (NAR) TTS framework, termed SimpleSpeech 2. SimpleSpeech 2 effectively combines the strengths of both autoregressive (AR) and non-autoregressive (NAR) methods, offering the following key advantages: (1) simplified data preparation; (2) straightforward model and loss design; and (3) stable, high-quality generation performance with fast inference speed. Compared to our previous publication, we present ({\romannumeral1}) a detailed analysis of the influence of speech tokenizer and noisy label for TTS performance; ({\romannumeral2}) four distinct types of sentence duration predictors; ({\romannumeral3}) a novel flow-based scalar latent transformer diffusion model. With these improvement, we show a significant improvement in generation performance and generation speed compared to our previous work and other state-of-the-art (SOTA) large-scale TTS models. Furthermore, we show that SimpleSpeech 2 can be seamlessly extended to multilingual TTS by training it on multilingual speech datasets. Demos are available on: {https://dongchaoyang.top/SimpleSpeech2\_demo/}.
Authors: Xinyang Hu, Fengzhuo Zhang, Siyu Chen, Zhuoran Yang
Abstract: Chain-of-Thought (CoT) prompting and its variants have gained popularity as effective methods for solving multi-step reasoning problems using pretrained large language models (LLMs). In this work, we analyze CoT prompting from a statistical estimation perspective, providing a comprehensive characterization of its sample complexity. To this end, we introduce a multi-step latent variable model that encapsulates the reasoning process, where the latent variable encodes the task information. Under this framework, we demonstrate that when the pretraining dataset is sufficiently large, the estimator formed by CoT prompting is equivalent to a Bayesian estimator. This estimator effectively solves the multi-step reasoning problem by aggregating a posterior distribution inferred from the demonstration examples in the prompt. Moreover, we prove that the statistical error of the CoT estimator can be decomposed into two main components: (i) a prompting error, which arises from inferring the true task using CoT prompts, and (ii) the statistical error of the pretrained LLM. We establish that, under appropriate assumptions, the prompting error decays exponentially to zero as the number of demonstrations increases. Additionally, we explicitly characterize the approximation and generalization errors of the pretrained LLM. Notably, we construct a transformer model that approximates the target distribution of the multi-step reasoning problem with an error that decreases exponentially in the number of transformer blocks. Our analysis extends to other variants of CoT, including Self-Consistent CoT, Tree-of-Thought, and Selection-Inference, offering a broad perspective on the efficacy of these methods. We also provide numerical experiments to validate the theoretical findings.
Authors: Shusaku Egami, Takahiro Ugai, Swe Nwe Nwe Htun, Ken Fukuda
Abstract: Multi-modal knowledge graphs (MMKGs), which ground various non-symbolic data (e.g., images and videos) into symbols, have attracted attention as resources enabling knowledge processing and machine learning across modalities. However, the construction of MMKGs for videos consisting of multiple events, such as daily activities, is still in the early stages. In this paper, we construct an MMKG based on synchronized multi-view simulated videos of daily activities. Besides representing the content of daily life videos as event-centric knowledge, our MMKG also includes frame-by-frame fine-grained changes, such as bounding boxes within video frames. In addition, we provide support tools for querying our MMKG. As an application example, we demonstrate that our MMKG facilitates benchmarking vision-language models by providing the necessary vision-language datasets for a tailored task.