Conference on Empirical Methods in Natural Language Processing · 593 papers
PunMemeCN: A Benchmark to Explore Vision-Language Models’ Understanding of Chinese Pun Memes
Zhijun Xu (Fudan University), Deqing Yang (Fudan University)
CodeClassificationGenerationTransformerSupervised Fine-TuningPrompt EngineeringVision Language ModelMultimodalityBenchmarkChain-of-Thought
🎯 What it does: Proposed and implemented the PUNMEMECN benchmark for evaluating the identification, sentiment analysis, and appropriate responses in chat scenarios for Chinese pun memes.
QFrCoLA: a Quebec-French Corpus of Linguistic Acceptability Judgments
David Beauchemin (Laval University), Richard Khoury (Laval University)
CodeClassificationData-Centric LearningTransformerLarge Language ModelSupervised Fine-TuningTextBenchmark
🎯 What it does: Proposed and released QFrCoLA, a binary dataset for French acceptability judgment containing 25,153 in-domain examples and 2,675 out-of-domain examples, and conducted fine-tuning and zero-shot evaluation on various monolingual and cross-lingual Transformers using this dataset.
Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
Yao Fu (Case Western Reserve University), Pan Li (Case Western Reserve University)
CodeExplainability and InterpretabilityComputational EfficiencyTransformerLarge Language ModelPrompt EngineeringText
🎯 What it does: Proposed the TruthfulnessEval framework, systematically evaluating the authenticity of quantized large language models across three dimensions: logical reasoning, common sense judgment, and imitative false answers, while investigating prompt sensitivity and internal representations.
🎯 What it does: Proposed a new query-focused retrieval head QRHEAD and built a retriever QRRETRIEVER based on it for long-context reasoning and paragraph re-ranking.
Debrup Das (University of Massachusetts Amherst), Razieh Rahimi (University of Massachusetts Amherst)
CodeRetrievalTransformerLarge Language ModelPrompt EngineeringContrastive LearningText
🎯 What it does: Built and trained a reasoning-oriented dense retriever called RaDeR, leveraging retrieval trajectories generated by MCTS to provide the retrieval model with diverse and high-quality training data.
RAED: Retrieval-Augmented Entity Description Generation for Emerging Entity Linking and Disambiguation
Karim Ghonim (Sapienza University of Rome), Roberto Navigli (Sapienza University of Rome)
CodeGenerationRetrievalTransformerLarge Language ModelTextRetrieval-Augmented Generation
🎯 What it does: RAED addresses the problem of new entity linking and disambiguation by retrieving external knowledge and generating entity titles and definitions.
🎯 What it does: Propose the RAG-Instruct method, which enhances the RAG capabilities of large language models by synthesizing diverse, high-quality retrieval-augmented instruction data, and constructs a 40K instruction set based on Wikipedia.
RAG-Zeval: Enhancing RAG Responses Evaluator through End-to-End Reasoning and Ranking-Based Reinforcement Learning
Kun Li (Chinese University of Hong Kong), Helen M. Meng
CodeRetrievalTransformerLarge Language ModelReinforcement LearningTextRetrieval-Augmented Generation
🎯 What it does: Built a rule-guided end-to-end evaluator, RAG-Zeval, for reliable and interpretable assessment of answers generated by retrieval-augmented generation (RAG) systems, and trained a compact LLM using reinforcement learning to achieve automated evaluation without human annotations.
RAGferee: Building Contextual Reward Models for Retrieval-Augmented Generation
Andrei C. Coman (Idiap Research Institute), Adrià de Gispert (Amazon AGI)
CodeData-Centric LearningReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningContrastive LearningTextRetrieval-Augmented Generation
🎯 What it does: Propose the RAGferee method, which retransforms QA datasets into RAG-specific preference pairs for training contextual reward models.
RALS: Resources and Baselines for Romanian Automatic Lexical Simplification
Fabian Anghel, Sergiu Nisioi (University of Bucharest)
CodeGenerationTransformerLarge Language ModelTextBenchmark
🎯 What it does: Created a dataset for predicting and simplifying lexical complexity in Romanian, and proposed a hybrid simplification system based on DexFlex.
Rapid Word Learning Through Meta In-Context Learning
Wentao Wang (New York University), Brenden Lake
CodeMeta LearningTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Proposes the Minnow method, leveraging meta-learning and placeholder tokens to enable language models to rapidly learn new words and generate novel usages with few examples;
RAVEN: Query-Guided Representation Alignment for Question Answering over Audio, Video, Embedded Sensors, and Natural Language
Subrata Biswas (Worcester Polytechnic Institute), Bashima Islam (Worcester Polytechnic Institute)
CodeRepresentation LearningTransformerLarge Language ModelSupervised Fine-TuningContrastive LearningVideoTextMultimodalityBenchmarkAudio
🎯 What it does: Propose the RAVEN framework, which adaptively fuses video, audio, sensor, and language modalities through query-guided cross-modal gating (QuART) to achieve multi-modal question answering.
Real-time Ad Retrieval via LLM-generative Commercial Intention for Sponsored Search Advertising
Tongtong Liu (Tencent Inc.), Peng Shu (Tencent Inc.)
CodeRetrievalRecommendation SystemTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Propose a real-time ad retrieval framework RARE based on LLM, which directly matches user queries with ads by utilizing commercial intent (CI) text generated by LLM.
Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills
Changsheng Wang (Michigan State University), Sijia Liu (Michigan State University)
CodeSafty and PrivacyTransformerTextChain-of-Thought
🎯 What it does: Studied the machine forgetting problem in Large Retrieval Models (LRM), proposing a new method R2MU that can completely remove sensitive information (including reasoning trajectories and final answers) without compromising inference capabilities.
RECALL: REpresentation-aligned Catastrophic-forgetting ALLeviation via Hierarchical Model Merging
Bowen Wang (Tsinghua University), Sheng Zhang (Tsinghua University)
CodeKnowledge DistillationRepresentation LearningTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Propose a hierarchical model fusion framework called RECALL based on internal representation similarity, aiming to achieve continual learning and catastrophic forgetting suppression for LLMs without relying on historical data or task labels.
🎯 What it does: Construct and pretrain a foundational model named RecBase, leveraging unified text representations and hierarchical concept IDs for autoregressive training to achieve zero-shot cross-domain recommendation.
RecGPT: A Foundation Model for Sequential Recommendation
Yangqin Jiang (University of Hong Kong), Chao Huang (University of Hong Kong)
CodeRecommendation SystemTransformerLarge Language ModelTextSequential
🎯 What it does: Propose a foundational model for sequence recommendation called RecGPT, which utilizes text-driven unified tokenization and autoregressive modeling to achieve zero-shot cross-domain recommendation;
Refining Attention for Explainable and Noise-Robust Fact-Checking with Transformers
Jean-Flavien Bussotti (Megagon Labs), Paolo Papotti (EURECOM)
CodeClassificationExplainability and InterpretabilityTransformerText
🎯 What it does: Propose the ATTUN model, which enhances model interpretability and robustness to noise by directly adjusting attention weights within the Transformer architecture.
Yifan Xia (Wuhan University), Jindong Gu (University of Oxford)
CodeOptimizationSafty and PrivacyLarge Language ModelSupervised Fine-TuningPrompt EngineeringVision Language ModelImageMultimodality
🎯 What it does: Proposes a method that simultaneously alleviates over-rejection and security issues against jailbreaking attacks in multi-modal large language models (MLLMs) by optimizing visual inputs (Magic Image).
Reinforced Query Reasoners for Reasoning-intensive Retrieval Tasks
Xubo Qin (State Key Laboratory of General Artificial Intelligence), Zilong Zheng (State Key Laboratory of General Artificial Intelligence)
CodeRetrievalTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: Propose and train a small-scale language model (TongSearch QR 7B/1.5B) for query reasoning and rewriting, using reinforcement learning to enhance performance in reasoning retrieval tasks.
CodeData-Centric LearningAI Code AssistantLarge Language ModelAgentic AIPrompt EngineeringVision Language ModelGraphTabularRetrieval-Augmented Generation
🎯 What it does: Automatically construct knowledge graphs and utilize retrieval-augmented generation to create executable exploratory data analysis (EDA) notebooks, combining LLM with self-correcting agents to achieve high reliability and low-cost code generation.
ReMedy: Learning Machine Translation Evaluation from Human Preferences with Reward Modeling
Shaomu Tan (University of Amsterdam), Christof Monz (University of Amsterdam)
CodeReinforcement Learning from Human FeedbackTransformerLarge Language ModelSupervised Fine-TuningReinforcement LearningText
🎯 What it does: Propose ReMedy, a reward modeling-based MT evaluation framework that trains models using contrastive human preference data to assess translation quality.
🎯 What it does: This review systematically organizes and evaluates the 'representation potential' of foundation models in unimodal and cross-modal alignment, including visual, language, speech, multimodal, and alignment phenomena with neuroscience.
Rescorla-Wagner Steering of LLMs for Undesired Behaviors over Disproportionate Inappropriate Context
Rushi Wang (University of Illinois Urbana-Champaign), Denghui Zhang (Stevens Institute of Technology)
CodeOptimizationSafty and PrivacyExplainability and InterpretabilityTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: This paper proposes a novel robust regulation method called RW-Steering, aiming to enhance the answer quality and safety of large language models in mixed contexts (containing both compliant and non-compliant information).
ReSeeding Latent States for Sequential Language Understanding
Stéphane Aroca-Ouellette (University of Colorado Boulder), Alessandro Roncone (University of Colorado Boulder)
CodeRepresentation LearningTransformerLarge Language ModelAuto EncoderContrastive LearningTextSequential
🎯 What it does: Build and train a framework called RESEED, enabling large language models to learn from environmental states during the training phase and re-inject these states' latent representations during the inference phase to achieve better multi-step sequential reasoning.
🎯 What it does: Proposes the ReSURE framework, dynamically downscaling unreliable supervision signals in multi-round dialogue fine-tuning to enhance model robustness and response quality.
CodeAI Code AssistantTransformerLarge Language ModelTextBenchmark
🎯 What it does: Proposed a framework called RETHINKMCTS for code generation that employs Monte Carlo Tree Search (MCTS) to explore and continuously refine ideas.
Reverse Prompt Engineering: A Zero-Shot, Genetic Algorithm Approach to Language Model Inversion
Hanqing Li (Northwestern University), Diego Klabjan (Northwestern University)
CodeOptimizationTransformerLarge Language ModelPrompt EngineeringText
🎯 What it does: Propose an untrained reverse prompting engineering (RPE) framework under black-box, zero-shot, and output-restricted scenarios, leveraging the language model's inherent reasoning ability and genetic algorithm iteration optimization to recover the original prompt based on only 5 output texts.
ReviewRL: Towards Automated Scientific Review with RL
Sihang Zeng (University of Washington), Bowen Zhou (Tsinghua University)
CodeReinforcement Learning from Human FeedbackTransformerSupervised Fine-TuningReinforcement LearningTextReview/Survey PaperRetrieval-Augmented GenerationChain-of-Thought
🎯 What it does: Proposed the ReviewRL framework for automatically generating scientific paper review reports and providing paper scores.
Reward-Shifted Speculative Sampling Is An Efficient Test-Time Weak-to-Strong Aligner
Bolian Li (Purdue University), Ruqi Zhang (Purdue University)
CodeReinforcement Learning from Human FeedbackTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Propose a reward-offset speculative sampling (SSS) algorithm that achieves weak-to-strong alignment during inference using an unaligned target model and an aligned small draft model, enabling high-quality generation without external reward models.
RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data Synthesis
Jianwei Wang (South China University of Technology), Ziqian Zeng (South China University of Technology)
CodeData SynthesisSafty and PrivacyTransformerLarge Language ModelSupervised Fine-TuningReinforcement LearningTextFinance Related
🎯 What it does: Propose the RewardDS framework, which generates high-quality synthetic data through reward-driven data synthesis and self-optimization for privacy-preserving LLM fine-tuning.
RICO: Improving Accuracy and Completeness in Image Recaptioning via Visual Reconstruction
Yuchi Wang (Peking University), Xu Sun (Kuaishou Technology)
CodeGenerationTransformerLarge Language ModelReinforcement LearningVision Language ModelImageTextMultimodalityBenchmark
🎯 What it does: Developed a vision-reconstruction-based image recaptioning framework called RICO, which generates reference images using a text-to-image model and iteratively refines captions by comparing the original image with the reconstructed image via a large model; to reduce computational costs, a single-step version called RICO-Flash was also introduced through Direct Preference Optimization (DPO) learning.
🎯 What it does: This paper addresses the insufficient modeling of audio event relationships in text-to-audio (TTA) generation by constructing a relation corpus containing 11 relationships and 25 audio event classes, and proposing a multi-stage relation-aware evaluation metric. Subsequently, a Gated Prompt Tuning strategy is introduced, significantly enhancing the relationship modeling performance of existing TTA models while keeping the parameter increment extremely low.
Robust Adaptation of Large Multimodal Models for Retrieval Augmented Hateful Meme Detection
Jingbiao Mei (University of Cambridge), Bill Byrne (University of Cambridge)
CodeClassificationDomain AdaptationExplainability and InterpretabilityTransformerLarge Language ModelSupervised Fine-TuningVision Language ModelContrastive LearningMultimodalityRetrieval-Augmented Generation
🎯 What it does: Propose the RA-HMD framework, which fine-tunes large multimodal models (LMM) using a two-stage LoRA + contrastive learning approach, combined with a retrieval-enhanced KNN classifier, to improve the accuracy, robustness, and interpretability of hate meme detection.
Robust Native Language Identification through Agentic Decomposition
Ahmet Yavuz Uluslu (University of Zurich), Rico Sennrich (University of Zurich)
CodeClassificationTransformerLarge Language ModelAgentic AIPrompt EngineeringText
🎯 What it does: Propose a NLI process based on agent splitting, where specialized agents first identify various linguistic features such as syntax, lexicon, and idioms, and then a coordinator agent synthesizes this evidence to make the final native language judgment.
RTQA : Recursive Thinking for Complex Temporal Knowledge Graph Question Answering with Large Language Models
Zhaoyan Gong (Zhejiang University), Wen Zhang (Zhejiang University)
CodeTransformerLarge Language ModelTextGraphRetrieval-Augmented GenerationChain-of-Thought
🎯 What it does: Developed a training-free, plug-and-play framework called RTQA, which recursively decomposes complex temporal knowledge graph question answering problems and performs bottom-up reasoning using large language models (LLMs) and temporal knowledge graphs.
CodeComputational EfficiencyData-Centric LearningTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: For reasoning tasks on language models, an extremely simple test-time scaling method called 'budget forcing' is proposed, along with the construction of a high-quality reasoning dataset s1K containing only 1,000 questions. Supervised fine-tuning of Qwen2.5-32B-Instruct on this dataset yields the model s1-32B, which can achieve a performance improvement trend with increased test-time computation by controlling the length of the thinking process (maximum token count or multiple additions of 'Wait').
SABER: Uncovering Vulnerabilities in Safety Alignment via Cross-Layer Residual Connection
Maithili Joshi (Indian Institute of Technology Delhi), Tanmoy Chakraborty (Indian Institute of Technology Delhi)
CodeSafty and PrivacyTransformerLarge Language ModelTextBenchmark
🎯 What it does: The study proposes a white-box jailbreak method called SABER, which bypasses the safety alignment mechanism of LLMs by inserting cross-layer residual connections between two layers.
SAEs Are Good for Steering – If You Select the Right Features
Dana Arad (Technion Israel Institute of Technology), Yonatan Belinkov
CodeExplainability and InterpretabilityRepresentation LearningAdversarial AttackAuto EncoderTextBenchmark
🎯 What it does: This paper studies how to select the most effective steering features by analyzing input/output scores of sparse autoencoder (SAE) features to achieve unsupervised control over language model generation.
SAFENUDGE: Safeguarding Large Language Models in Real-time with Tunable Safety-Performance Trade-offs
Joao Fonseca (New York University), Julia Stoyanovich (New York University)
CodeSafty and PrivacyLarge Language ModelText
🎯 What it does: Propose SAFENUDGE, a real-time safety protection mechanism for text generation, which suppresses jailbreak attacks by controlling text generation and guidance.
SAKI-RAG: Mitigating Context Fragmentation in Long-Document RAG via Sentence-level Attention Knowledge Integration
Wenyu Tao (South China University of Technology), Xiangmin Xu (South China University of Technology)
CodeRetrievalTransformerLarge Language ModelTextRetrieval-Augmented Generation
🎯 What it does: Propose the SAKI-RAG framework, utilizing sentence-level attention linking and dual-axis retrieval technology to address the context fragmentation problem in long document RAG.
Sali4Vid: Saliency-Aware Video Reweighting and Adaptive Caption Retrieval for Dense Video Captioning
MinJu Jeon (Hanyang University), Dong-Jin Kim (Hanyang University)
CodeGenerationRetrievalTransformerVision Language ModelVideoTextMultimodalityRetrieval-Augmented Generation
🎯 What it does: Proposed the Sali4Vid framework, combining significance video reweighting and semantic adaptive caption retrieval to achieve dense video caption generation and event localization.
Scaling Low-Resource MT via Synthetic Data Generation with LLMs
Ona de Gibert (University of Helsinki), Jörg Tiedemann (University of Cambridge)
CodeData SynthesisTransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Leveraged GPT-4o for large-scale forward translation to construct a document-level synthetic parallel corpus containing 7 low-resource languages, and extended it to 147 language pairs through a pivot mechanism; conducted automatic and manual quality assessments of the corpus, trained and fine-tuned multiple MT models on it, and ultimately released the public synthetic corpus SynOPUS.
Scaling Up Temporal Domain Generalization via Temporal Experts Averaging
Aoming Liu (Boston University), Bryan A. Plummer (Boston University)
CodeDomain AdaptationConvolutional Neural NetworkTransformerSupervised Fine-TuningMixture of ExpertsImageTextBenchmark
🎯 What it does: Propose a framework called Temporal Experts Averaging (TEA) that uses full-model weight averaging to address the Temporal Domain Generalization (TDG) problem.
Bofu Dong (Indiana University), Ming Jiang (Indiana University)
CodeLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextBenchmark
🎯 What it does: Propose the SciEvent benchmark, which annotates multi-domain scientific abstracts using a unified event extraction (EE) scheme, covering five disciplines.
SciNLP: A Domain-Specific Benchmark for Full-Text Scientific Entity and Relation Extraction in NLP
Decheng Duan (Nanjing University of Science and Technology), Chengzhi Zhang (Nanjing University of Science and Technology)
CodeRecognitionTransformerLarge Language ModelTextGraphBenchmark
🎯 What it does: Constructed SciNLP, the first fine-grained entity and relation extraction benchmark dataset tailored for the NLP domain, covering full-text documents.
SciRIFF: A Resource to Enhance Language Model Instruction-Following over Scientific Literature
David Wadden (Allen Institute for AI), Arman Cohan (Yale University)
CodeClassificationLarge Language ModelSupervised Fine-TuningTextBiomedical DataBenchmark
🎯 What it does: This paper constructs SciRIFF, a science literature-oriented instruction-following dataset containing 137K expert-written instructions, covering 54 tasks across five capabilities: information extraction, summarization, question answering, statement verification, and classification.
SCRIBE: Structured Chain Reasoning for Interactive Behaviour Explanations using Tool Calling
Fares Fawzi (École Polytechnique Fédérale de Lausanne), Tanja Käser (École Polytechnique Fédérale de Lausanne)
CodeData SynthesisExplainability and InterpretabilityTransformerLarge Language ModelSupervised Fine-TuningAgentic AITextChain-of-Thought
🎯 What it does: Propose the SCRIBE framework, enabling small language models to generate personalized, verifiable student behavior feedback explanations through multi-hop tool calls;
CodeReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: This paper proposes a self-reinforcement learning framework (SDGO) based on the discriminative capabilities of LLMs, achieving enhanced generation safety during training by using the model's safety discrimination results during the generation phase as rewards;
SEA: Supervised Embedding Alignment for Token-Level Visual-Textual Integration in MLLMs
Yuanyang Yin (MoE Key Lab of BIPC University of Science and Technology of China), Feng Zhao (MoE Key Lab of BIPC University of Science and Technology of China)
CodeRepresentation LearningTransformerSupervised Fine-TuningVision Language ModelContrastive LearningMultimodalityBenchmark
🎯 What it does: Enhancing the alignment quality between vision and text in the LLM embedding space through the SEA method, which introduces token-level supervision during the pre-training phase of multi-modal large language models.
🎯 What it does: Propose the SEAL framework, leveraging structure-aware contrastive learning and element-level alignment to enhance long-structured document retrieval; and release the StructDocRetrieval dataset with structural annotations.
Searching for the Most Human-like Emergent Language
Brendon Boldt (Carnegie Mellon University), David R. Mortensen (Carnegie Mellon University)
CodeGenerationHyperparameter SearchTextBenchmark
🎯 What it does: Generate emergent languages with the highest similarity to human language statistics by performing hyperparameter search in a signaling game environment, and evaluate them on XferBench deep transfer learning tasks.
CodeSegmentationVision Language ModelMultimodalityBenchmark
🎯 What it does: Proposed the Seeing Culture Benchmark (SCB), evaluating cultural reasoning abilities of multimodal vision-language models through two-stage visual question answering and cultural object segmentation
Seeing the Same Story Differently: Framing‐Divergent Event Coreference for Computational Framing Analysis
Jin Zhao (Brandeis University), Nianwen Xue (Brandeis University)
CodeClassificationTransformerLarge Language ModelSupervised Fine-TuningReinforcement LearningTextBenchmark
🎯 What it does: Proposed and implemented the FRECO task, which identifies event pairs presenting the same real-world event with contrasting frameworks in different news texts.
Self-Augmented Preference Alignment for Sycophancy Reduction in LLMs
Chien Hung Chen, Hsin-Hsi Chen (Academia Sinica)
CodeData-Centric LearningReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: Investigate the 'flattery bias' in open-source large language models during the alignment process, and propose an alignment method that reduces this bias by having the model evaluate user suggestions.
Self-Critique and Refinement for Faithful Natural Language Explanations
Yingming Wang (University of Copenhagen), Pepa Atanasova (University of Copenhagen)
CodeExplainability and InterpretabilityTransformerLarge Language ModelPrompt EngineeringText
🎯 What it does: Enhance the credibility of natural language explanations (NLE) generated by models through an SR-NLE framework that realizes self-criticism and iterative improvement within a single LLM.
SelfRACG: Enabling LLMs to Self-Express and Retrieve for Code Generation
Qian Dong (Tsinghua University), Shaoping Ma (Tsinghua University)
CodeGenerationRetrievalAI Code AssistantLarge Language ModelContrastive LearningTextBenchmarkRetrieval-Augmented Generation
🎯 What it does: Propose the SelfRACG framework, enabling LLMs to self-express information needs, thereby enhancing generation quality in retrieval-augmented code generation;
SEMMA: A Semantic Aware Knowledge Graph Foundation Model
Arvindh Arun (Institute for AI University of Stuttgart), Steffen Staab
CodeRepresentation LearningGraph Neural NetworkLarge Language ModelTextGraph
🎯 What it does: Propose the SEMMA model, which achieves zero-shot link prediction by leveraging dual-modal information generated by LLMs, including relational text semantics and graph structure.
CodeAutonomous DrivingGraph Neural NetworkTransformerLarge Language ModelAgentic AITextMultimodalityTime Series
🎯 What it does: Propose the SeMob framework, which automatically extracts event texts using multi-agent LLMs and fuses them with spatiotemporal traffic data to enhance event-driven urban mobility prediction.
Sentence Smith: Controllable Edits for Evaluating Text Embeddings
Hongji Li (University of Zurich), Juri Opitz (University of Zurich)
CodeExplainability and InterpretabilityRepresentation LearningTransformerLarge Language ModelPrompt EngineeringContrastive LearningTextGraphBenchmark
🎯 What it does: This paper proposes the SENTENCESMITH framework, which utilizes semantic graphs (AMR) for controllable and transparent transformation of sentence meaning, and applies this framework to generate challenging negative samples for text embedding models, thereby achieving fine-grained model evaluation.
Separate the Wheat from the Chaff: Winnowing Down Divergent Views in Retrieval Augmented Generation
Song Wang (University of Virginia), Jundong Li (University of Virginia)
CodeGenerationRetrievalTransformerLarge Language ModelTextRetrieval-Augmented Generation
🎯 What it does: Propose the WinnowRAG framework, which groups retrieved documents into perspective-based clusters through query-aware clustering, and iteratively filters noise documents while preserving valuable information via a multi-agent winnowing stage with a Critic LLM, ultimately improving the accuracy of retrieval-augmented generation.
SEPS: A Separability Measure for Robust Unlearning in LLMs
Wonje Jeung (Yonsei University), Albert No (Yonsei University)
CodeTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextBenchmark
🎯 What it does: This paper investigates the robustness of machine unlearning in large language models, proposing a new evaluation metric called SEPS to measure the model's ability to separately handle forgetting (forget) and retaining (retain) queries within the same prompt; after discovering that existing methods fail under mixed prompts, we propose a Mixed Prompt (MP) training strategy, developing two approaches, MP-ME (target-free) and MP-IDK (targeted), to enhance the model's forgetting and retaining performance in mixed query scenarios.
Sheaf Discovery with Joint Computation Graph Pruning and Flexible Granularity
Lei Yu (University Of Toronto), Gerald Penn (University Of Toronto)
CodeComputational EfficiencyTransformerLarge Language ModelText
🎯 What it does: Proposed the concept of 'sheaf' and automatically discovered self-contained modules capable of executing specific tasks individually within pre-trained language models through the DiscoGP framework.
SHIFT: Selected Helpful Informative Frame for Video-guided Machine Translation
Boyu Guan (Chinese Academy of Sciences), Chengqing Zong (Chinese Academy of Sciences)
CodeConvolutional Neural NetworkTransformerLarge Language ModelVision Language ModelVideoTextMultimodality
🎯 What it does: Propose the SHIFT framework for video-assisted machine translation, dynamically deciding whether to use video, selecting only one keyframe or using text as input, and adapting to multimodal large language models.
Should I Share this Translation? Evaluating Quality Feedback for User Reliance on Machine Translation
Dayeon Ki (University of Maryland), Marine Carpuat (University of Maryland)
CodeExplainability and InterpretabilityTransformerLarge Language ModelText
🎯 What it does: This paper evaluates the impact of four quality feedback mechanisms (error highlighting, LLM explanations, back-translation, and Q&A tables) on monolingual users' decision accuracy and appropriate reliance when deciding whether to share translations in real-world machine translation scenarios;
🎯 What it does: Studies how to leverage user-AI assistant consultation logs to enhance personalized search effectiveness on e-commerce platforms, proposing a search model called VAPS based on consultation value assessment.
SimpleDoc: Multi‐Modal Document Understanding with Dual‐Cue Page Retrieval and Iterative Refinement
Chelsi Jain (Oregon State University), Huazheng Wang (Oregon State University)
CodeRetrievalComputational EfficiencyTransformerLarge Language ModelVision Language ModelMultimodalityBenchmarkRetrieval-Augmented GenerationChain-of-Thought
🎯 What it does: Propose the SimpleDoc framework to achieve multi-modal document visual question answering by generating answers through dual-clue retrieval and iterative reasoning.
🎯 What it does: Propose a unified retrieval-augmented generation framework called RoleRAG, which enables a single frozen LLM to perform multi-tasking by utilizing role-specific tokens;
🎯 What it does: Propose the Sketch-of-Thought (SoT) framework, which compresses LLM reasoning processes by leveraging three cognitive-inspired reasoning paradigms (concept chains, chunked symbols, expert dictionaries) and dynamically selects paradigms through a lightweight router
SLlama: Parameter-Efficient Language Model Architecture for Enhanced Linguistic Competence Under Strict Data Constraints
Victor Adelakun Omolaoye (University of Potsdam), Gerard de Melo (University of Potsdam)
CodeComputational EfficiencyRepresentation LearningTransformerLarge Language ModelTextBenchmark
🎯 What it does: This paper proposes SLlama, a compact variant of Llama-3 trained on the BabyLM dataset with only 10 million tokens without distillation, demonstrating outstanding linguistic capabilities;
SLoW: Select Low-frequency Words! Automatic Dictionary Selection for Translation on Large Language Models
Hongyuan Lu (Chinese University of Hong Kong), Wai Lam (Jilin University)
CodeGenerationComputational EfficiencyTransformerLarge Language ModelPrompt EngineeringText
🎯 What it does: Proposed the Automatic Dictionary Selection (ADS) task and developed the SLoW method, which selects dictionaries based on low-frequency words to save tokens and enhance performance in large language model (LLM) translation.
SNaRe: Domain-aware Data Generation for Low-Resource Event Detection
Tanmay Parekh (University of California Los Angeles), Nanyun Peng (University of California Los Angeles)
CodeGenerationData SynthesisDomain AdaptationTransformerLarge Language ModelPrompt EngineeringText
🎯 What it does: In low-resource scenarios for event detection, the SNARE framework is proposed, which generates high-quality synthetic data using unlabeled target domain data;
so much depends / upon / a whitespace: Why Whitespace Matters for Poets and LLMs
Sriharsh Bhyravajjula (University of Washington), Maria Antoniak (University of Colorado Boulder)
CodeClassificationLarge Language ModelVision-Language-Action ModelTextMultimodalityBenchmark
🎯 What it does: This paper introduces WISP fine-grained whitespace symbol classification, systematically analyzing and comparing whitespace usage patterns in 19,000 publicly available poems, 12,000 Reddit unpublished poems, and 51,000 LLM-generated poems.
Social Bias in Multilingual Language Models: A Survey
Lance Calvin Lim Gamboa (University of Birmingham), Mark G. Lee (University of Birmingham)
CodeTransformerReview/Survey Paper
🎯 What it does: Systematic review of social bias in multilingual models, analyzing language diversity, cultural awareness, and evaluation and mitigation methods.
SOCIAL SCAFFOLDS: A Generalization Framework for Social Understanding Tasks
Ritam Dutt (Carnegie Mellon University), Maarten Sap (Carnegie Mellon University)
CodeClassificationTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringText
🎯 What it does: Propose the SOCIAL SCAFFOLDS framework, which enhances the generalization ability of multi-task dialogue understanding models by utilizing automatically generated social reasoning 'rationales' (intentions, listener responses, presuppositions).
SocioBench: Modeling Human Behavior in Sociological Surveys with Large Language Models
Jia Wang (Shanghai Innovation Institute), Zhongyu Wei (Fudan University)
CodeLarge Language ModelPrompt EngineeringTextTabularBenchmark
🎯 What it does: Propose the cross-national and cross-domain SocioBench benchmark, using large language models to simulate human behavior in social surveys and assessing the alignment between the simulated behavior and real survey results.
SOLAR: Towards Characterizing Subjectivity of Individuals through Modeling Value Conflicts and Trade-offs
Younghun Lee (Purdue University), Dan Goldwasser (Purdue University)
CodeTransformerLarge Language ModelTextRetrieval-Augmented Generation
🎯 What it does: Built a retrieval-augmented generation framework called SOLAR, which infers individuals' subjective judgments by leveraging their past Reddit comments.
CodeReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningTextMultimodalityChain-of-ThoughtAudio
🎯 What it does: Propose the SoundMind dataset and the SoundMind-RL algorithm for training and evaluating the logical reasoning capabilities of audio-language models.
SPARK: Simulating the Co-evolution of Stance and Topic Dynamics in Online Discourse with LLM-based Agents
Bowen Zhang (Shenzhen Technology University), Hu Huang (University of Science and Technology of China)
CodeExplainability and InterpretabilityTransformerLarge Language ModelAgentic AIText
🎯 What it does: Proposes a multi-agent simulation framework named SPARK based on large language models (LLMs) to jointly model the co-evolutionary process of topic evolution and stance changes in online discussions.
Sparse Autoencoder Features for Classifications and Transferability
Jack Gallifant (Harvard University), Danielle Bitterman
CodeClassificationDomain AdaptationRepresentation LearningLarge Language ModelAuto EncoderImageTextMultimodality
🎯 What it does: Investigate the interpretable classification and cross-modal transfer of sparse autoencoder (SAE) features in large language models, systematically evaluating the impact of model layers, width, pooling, and binarization on performance.
🎯 What it does: Constructed a large-scale news homepage dataset called NewsHomepages, and used the spatial layout (position, size, image) of the homepage as weak labels to train a Transformer model for predicting editors' prioritization of news items.
CodeComputational EfficiencyRepresentation LearningData-Centric LearningTransformerVision Language ModelContrastive LearningMultimodality
🎯 What it does: Proposed a no-reference evaluation metric for long image captions called SPECS, based on CLIP and incorporating a training objective specifically tailored for details.
SportReason: Evaluating Retrieval-Augmented Reasoning across Tables and Text for Sports Question Answering
Kaiyue Feng (New York University Shanghai), Chen Zhao (Center for Data Science New York University)
CodeRetrievalLarge Language ModelAgentic AIMultimodalityTabularBenchmarkRetrieval-Augmented Generation
🎯 What it does: Introduce the SPORTREASON benchmark to evaluate the effectiveness of retrieval-augmented reasoning across tables and text in sports question answering.
SQLWOZ: A Realistic Task-Oriented Dialogue Dataset with SQL-Based Dialogue State Representation for Complex User Requirements
Heng-Da Xu (Beijing Institute of Technology), Heyan Huang (Beijing Institute of Technology)
CodeTransformerLarge Language ModelTextBenchmark
🎯 What it does: Constructed SQLWOZ—a multi-domain task-oriented dialog dataset that represents dialog states using SQL statements, capable of reflecting multi-value, exclusion, priority, and conditional constraint slots, and conducted dialog state tracking, generation, and end-to-end task-oriented dialog experiments on this dataset.
SQUiD: Synthesizing Relational Databases from Unstructured Text
Mushtari Sadia (University of Michigan), Amrita Roy Chowdhury (University of Michigan)
CodeData SynthesisTransformerLarge Language ModelPrompt EngineeringTextTabularChain-of-Thought
🎯 What it does: This study proposes the SQUiD framework to automatically synthesize structured relational databases from unstructured text, encompassing four stages: schema design, value extraction, table filling, and SQL generation;
SSA: Semantic Contamination of LLM-Driven Fake News Detection
Cheng Xu (University College Dublin), Tahar Kechadi (University College Dublin)
CodeClassificationExplainability and InterpretabilityTransformerLarge Language ModelPrompt EngineeringTextBenchmark
🎯 What it does: Proposed an evaluation framework named Semantic Sensitivity Amplifier (SSA) to detect semantic-level benchmark data contamination (BDC) in large language models for fake news detection tasks, assessing model robustness through entity replacement perturbation and the SSA Factor metric.
STARQA: A Question Answering Dataset for Complex Analytical Reasoning over Structured Databases
Mounica Maddela (Bloomberg), Mausam (Indian Institute of Technology Delhi)
CodeAI Code AssistantLarge Language ModelTabularBenchmark
🎯 What it does: Proposed the STARQA public dataset and designed the TEXT2SQLCODE method that integrates SQL with Python, evaluating the performance of LLMs in complex analytical reasoning question answering.
Chengpeng Li (University of Science and Technology of China), Dayiheng Liu (Qwen Team, Alibaba Inc)
CodeAI Code AssistantTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextChain-of-Thought
🎯 What it does: Design the START framework, which activates the tool usage capability of LLMs through untrained Hint-infer, and further enhances the model's utilization of tools via self-supervised Hint-RFT (rejection sampling), thereby achieving more accurate reasoning and a more efficient thinking process.
Static or Dynamic: Towards Query-Adaptive Token Selection for Video Question Answering
Yumeng Shi (Nanyang Technological University), Wenya Wang (Nanyang Technological University)
CodeRetrievalTransformerVision Language ModelVideoTextMultimodality
🎯 What it does: Proposed a query-adaptive Token selection framework named EXPLORE-THEN-SELECT to dynamically balance static and dynamic information in video question answering under limited Token budget.
🎯 What it does: Propose a new static word embedding (SWE) method that extracts word vectors using a pre-trained sentence Transformer, then fine-tunes them via sentence-level PCA (with the first few principal components removed) and knowledge distillation or contrastive learning, resulting in a lightweight model where word vectors can be directly averaged into sentence vectors;
Steering Language Models in Multi-Token Generation: A Case Study on Tense and Aspect
Alina Klerings (University of Mannheim), Simone Paolo Ponzetto (University of Mannheim)
CodeGenerationExplainability and InterpretabilityRepresentation LearningTransformerLarge Language ModelText
🎯 What it does: Investigate the internal representations of tense and aspect in large language models (Llama-3.1-8B and Qwen-2.5-7B), and achieve syntactic control over multi-word generation through linear concept-driven interventions.
CodeComputational EfficiencyTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: This paper proposes a step-level verifier-guided hybrid test-time scaling method to enhance the reasoning performance of large language models.
CodeGenerationRetrievalTransformerLarge Language ModelReinforcement LearningTextRetrieval-Augmented Generation
🎯 What it does: Propose STEPSEARCH, a reinforcement learning framework based on token-level stepping rewards for training LLMs to reason through search interactions in multi-hop question answering.
🎯 What it does: Propose a model fusion method based on the linear representation hypothesis—STF, which achieves multi-task model fusion without training by performing singular value decomposition (SVD) on the task matrix and solving a linear system in the singular space.
SURE: Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models
Yuxin Gou (Hefei University of Technology), Wenbo Hu (Hefei University of Technology)
CodeSafty and PrivacySupervised Fine-TuningMultimodalityChain-of-Thought
🎯 What it does: Propose the SURE framework, training MLLM to achieve safe intent recognition and rejection for multi-modal inputs through chain-of-thought reasoning.