{"total":91,"rows":[{"id":"2403.05527","title":"GEAR: An Efficient KV Cache Compression Recipe for Near-Lossless Generative Inference of LLM","created":"2024-03-08","cited_by_count":158,"area":"inference"},{"id":"2503.03321","title":"See What You Are Told: Visual Attention Sink in Large Multimodal Models","created":"2025-03-05","cited_by_count":132,"area":"mllm"},{"id":"2402.01761","title":"Rethinking Interpretability in the Era of Large Language Models","created":"2024-01-30","cited_by_count":111,"area":"lr"},{"id":"2406.16008","title":"Found in the Middle: Calibrating Positional Attention Bias Improves Long Context Utilization","created":"2024-06-23","cited_by_count":101,"area":null},{"id":"2411.00348","title":"Attention Tracker: Detecting Prompt Injection Attacks in LLMs","created":"2024-11-01","cited_by_count":95,"area":"safety"},{"id":"2411.11296","title":"Steering Language Model Refusal with Sparse Autoencoders","created":"2024-11-18","cited_by_count":95,"area":"mechinterp"},{"id":"2503.01773","title":"Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areas","created":"2025-03-03","cited_by_count":95,"area":"lr"},{"id":"2406.15765","title":"Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibration","created":"2024-06-22","cited_by_count":74,"area":null},{"id":"2503.24370","title":"Effectively Controlling Reasoning Models through Thinking Intervention","created":"2025-03-31","cited_by_count":65,"area":"lr"},{"id":"2406.01563","title":"LoFiT: Localized Fine-tuning on LLM Representations","created":"2024-06-03","cited_by_count":57,"area":null},{"id":"2502.17516","title":"A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models","created":"2025-02-22","cited_by_count":34,"area":"mllm"},{"id":"2404.15846","title":"From Complex to Simple: Enhancing Multi-Constraint Complex Instruction Following Ability of Large Language Models","created":"2024-04-24","cited_by_count":27,"area":null},{"id":"2401.13986","title":"Towards Consistent Natural-Language Explanations via Explanation-Consistency Finetuning","created":"2024-01-25","cited_by_count":26,"area":null},{"id":"2410.14516","title":"Do LLMs \"know\" internally when they follow instructions?","created":"2024-10-18","cited_by_count":26,"area":null},{"id":"2410.14582","title":"Do LLMs estimate uncertainty well in instruction-following?","created":"2024-10-18","cited_by_count":26,"area":null},{"id":"2408.16119","title":"Data Formulator 2: Iterative Creation of Data Visualizations, with AI Transforming Data Along the Way","created":"2024-08-28","cited_by_count":21,"area":null},{"id":"2505.12025","title":"Spotlight Your Instructions: Instruction-following with Dynamic Attention Steering","created":"2025-05-17","cited_by_count":21,"area":null},{"id":"2406.19598","title":"Mixture of In-Context Experts Enhance LLMs' Long Context Awareness","created":"2024-06-28","cited_by_count":17,"area":null},{"id":"2409.10790","title":"Model Tells Itself Where to Attend: Faithfulness Meets Automatic Attention Steering","created":"2024-09-16","cited_by_count":17,"area":null},{"id":"2409.13761","title":"Do Large Language Models Need a Content Delivery Network?","created":"2024-09-16","cited_by_count":16,"area":"inference"},{"id":"2406.03631","title":"Discovering Bias in Latent Space: An Unsupervised Debiasing Approach","created":"2024-06-05","cited_by_count":15,"area":null},{"id":"2407.06426","title":"DebUnc: Improving Large Language Model Agent Communication With Uncertainty Metrics","created":"2024-07-08","cited_by_count":14,"area":"agents"},{"id":"2502.13870","title":"SPEX: Scaling Feature Interaction Explanations for LLMs","created":"2025-02-19","cited_by_count":12,"area":null},{"id":"2505.11739","title":"ZeroTuning: Unlocking the Initial Token's Power to Enhance Large Language Models Without Training","created":"2025-05-16","cited_by_count":12,"area":null},{"id":"2401.02575","title":"Large Language Models for Social Networks: Applications, Challenges, and Solutions","created":"2024-01-04","cited_by_count":11,"area":null},{"id":"2411.02688","title":"On the Loss of Context-awareness in General Instruction Fine-tuning","created":"2024-11-05","cited_by_count":11,"area":null},{"id":"2506.05451","title":"Interpretation Meets Safety: A Survey on Interpretation Methods and Tools for Improving LLM Safety","created":"2025-06-05","cited_by_count":10,"area":"safety"},{"id":"2512.08417","title":"Attention is All You Need to Defend Against Indirect Prompt Injection Attacks in LLMs","created":"2025-12-09","cited_by_count":10,"area":"agents"},{"id":"2602.20708","title":"ICON: Indirect Prompt Injection Defense for Agents based on Inference-Time Correction","created":"2026-02-24","cited_by_count":10,"area":"agents"},{"id":"2603.24543","title":"Analysing the Safety Pitfalls of Steering Vectors","created":"2026-03-25","cited_by_count":10,"area":"safety"},{"id":"2407.05578","title":"FALIP: Visual Prompt as Foveal Attention Boosts CLIP Zero-Shot Performance","created":"2024-07-08","cited_by_count":8,"area":null},{"id":"2602.00333","title":"Efficient and accurate steering of Large Language Models through attention-guided feature learning","created":"2026-01-30","cited_by_count":8,"area":null},{"id":"2603.01281","title":"Spectral Attention Steering for Prompt Highlighting","created":"2026-03-01","cited_by_count":8,"area":null},{"id":"2411.13009","title":"LLMSteer: Improving Long-Context LLM Inference by Steering Attention on Reused Contexts","created":"2024-11-20","cited_by_count":7,"area":null},{"id":"2502.15334","title":"Attention Eclipse: Manipulating Attention to Bypass LLM Safety-Alignment","created":"2025-02-21","cited_by_count":7,"area":"safety"},{"id":"2503.12908","title":"HICD: Hallucination-Inducing via Attention Dispersion for Contrastive Decoding to Mitigate Hallucinations in Large Language Models","created":"2025-03-17","cited_by_count":7,"area":null},{"id":"2410.22517","title":"Attention Speaks Volumes: Localizing and Mitigating Bias in Language Models","created":"2024-10-29","cited_by_count":6,"area":null},{"id":"2502.01925","title":"PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling","created":"2025-02-04","cited_by_count":6,"area":"alignment"},{"id":"2505.15291","title":"Hallucinate at the Last in Long Response Generation: A Case Study on Long Document Summarization","created":"2025-05-21","cited_by_count":6,"area":null},{"id":"2601.21742","title":"Epistemic Context Learning: Building Trust the Right Way in LLM-Based Multi-Agent Systems","created":"2026-01-29","cited_by_count":6,"area":"agents"},{"id":"2603.06745","title":"Enhancing Instruction Following of LLMs via Activation Steering with Dynamic Rejection","created":"2026-03-06","cited_by_count":6,"area":"mechinterp"},{"id":"2410.16327","title":"Feint and Attack: Attention-Based Strategies for Jailbreaking and Protecting LLMs","created":"2024-10-18","cited_by_count":5,"area":"safety"},{"id":"2510.22067","title":"Capturing Gaze Shifts for Guidance: Cross-Modal Fusion Enhancement for VLM Hallucination Mitigation","created":"2025-10-24","cited_by_count":5,"area":"mllm"},{"id":"2403.01002","title":"Attribute Structuring Improves LLM-Based Evaluation of Clinical Text Summaries","created":"2024-03-01","cited_by_count":4,"area":null},{"id":"2406.16779","title":"It Is Not About What You Say, It Is About How You Say It: A Surprisingly Simple Approach for Improving Reading Comprehension","created":"2024-06-24","cited_by_count":4,"area":null},{"id":"2502.15335","title":"Stepwise Informativeness Search for Efficient and Effective LLM Reasoning","created":"2025-02-21","cited_by_count":4,"area":"lr"},{"id":"2505.10063","title":"CAFE: Retrieval Head-based Coarse-to-Fine Information Seeking to Enhance Multi-Document QA Capability","created":"2025-05-15","cited_by_count":4,"area":null},{"id":"2411.00066","title":"Interpretable Next-token Prediction via the Generalized Induction Head","created":"2024-10-31","cited_by_count":3,"area":null},{"id":"2507.10085","title":"Enhancing Chain-of-Thought Reasoning with Critical Representation Fine-tuning","created":"2025-07-14","cited_by_count":3,"area":"lr"},{"id":"2511.05017","title":"Towards Mitigating Hallucinations in Large Vision-Language Models by Refining Textual Embeddings","created":"2025-11-07","cited_by_count":3,"area":"lr"},{"id":"2602.22719","title":"Interpreting and Steering State-Space Models via Activation Subspace Bottlenecks","created":"2026-02-26","cited_by_count":3,"area":"ssm"},{"id":"2507.12851","title":"Simulate, Refocus and Ensemble: An Attention-Refocusing Scheme for Domain Generalization","created":"2025-07-17","cited_by_count":2,"area":null},{"id":"2603.07837","title":"AI Steerability 360: A Toolkit for Steering Large Language Models","created":"2026-03-08","cited_by_count":2,"area":null},{"id":"2603.10705","title":"PRISM-$\\Delta$: Differential Subspace Steering for Prompt Highlighting in Large Language Models","created":"2026-03-11","cited_by_count":2,"area":null},{"id":"2604.06695","title":"Reasoning Fails Where Step Flow Breaks","created":"2026-04-08","cited_by_count":2,"area":"lr"},{"id":"2607.26228","title":"Steering Instruction Hierarchies at Inference Time","created":"2026-07-28","cited_by_count":2,"area":null},{"id":"2410.08469","title":"Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP","created":"2024-10-11","cited_by_count":1,"area":"mllm"},{"id":"2505.17855","title":"Explaining Sources of Uncertainty in Automated Fact-Checking","created":"2025-05-23","cited_by_count":1,"area":null},{"id":"2510.00586","title":"Eyes-on-Me: Scalable RAG Poisoning through Transferable Attention-Steering Attractors","created":"2025-10-01","cited_by_count":1,"area":"rag"},{"id":"2602.08882","title":"Designing Multi-Robot Ground Video Sensemaking with Public Safety Professionals","created":"2026-02-09","cited_by_count":1,"area":null},{"id":"2603.05772","title":"Depth Charge: Jailbreak Large Language Models from Deep Safety Attention Heads","created":"2026-03-06","cited_by_count":1,"area":"safety"},{"id":"2603.13359","title":"From Refusal Tokens to Refusal Control: Discovering and Steering Category-Specific Refusal Directions","created":"2026-03-09","cited_by_count":1,"area":"alignment"},{"id":"2603.18453","title":"Learning Consistent Temporal Grounding between Related Tasks in Sports Coaching","created":"2026-03-19","cited_by_count":1,"area":null},{"id":"2605.10622","title":"Vocabulary Hijacking in LVLMs: Unveiling Critical Attention Heads by Excluding Inert Tokens to Mitigate Hallucination","created":"2026-05-11","cited_by_count":1,"area":"mllm"},{"id":"2608.17095","title":"Inference-Time Attention Steering for Vision-Language-Action Driving Models","created":"2026-08-17","cited_by_count":1,"area":"vla"},{"id":"2608.24136","title":"Steering Recurrent Reasoners at Inference Time with Readout Feedback","created":"2026-08-25","cited_by_count":1,"area":null},{"id":"2609.02737","title":"Language Models Can Control Their Own Attention","created":"2026-09-02","cited_by_count":1,"area":"inference"},{"id":"2505.24292","title":"Mind the Quote: Enabling Quotation-Aware Dialogue in LLMs via Plug-and-Play Modules","created":"2025-05-30","cited_by_count":0,"area":null},{"id":"2506.14912","title":"CrEst: Credibility Estimation for Contexts in LLMs via Weak Supervision","created":"2025-06-17","cited_by_count":0,"area":null},{"id":"2605.23040","title":"Steered Generation via Gradient-Based Optimization on Sparse Query Features","created":"2026-05-21","cited_by_count":0,"area":"mechinterp"},{"id":"2605.30136","title":"Enhancing Multi-Agent Communication through Attention Steering with Context Relevance","created":"2026-05-28","cited_by_count":0,"area":"agents"},{"id":"2607.03738","title":"Attending to Multimodal Generation One Token at a Time","created":"2026-07-04","cited_by_count":0,"area":"mllm"},{"id":"2607.13395","title":"Self-Improving is Often Sudden: Enlightenment-style Finetuning for Large-Scale Models","created":"2026-07-15","cited_by_count":0,"area":"synth"},{"id":"2607.13425","title":"Data-Efficient Adaptation of LLMs via Attention Head Reweighting","created":"2026-07-15","cited_by_count":0,"area":null},{"id":"2607.25873","title":"How Do LLMs Read Bug Reports? An Empirical Study of Attention in LLMs for Automated Program Repair","created":"2026-07-28","cited_by_count":0,"area":null},{"id":"2608.03711","title":"Attention is Case-Sensitive","created":"2026-08-04","cited_by_count":0,"area":null},{"id":"2609.04721","title":"Locating and Steering Refusal Beyond Attention","created":"2026-09-04","cited_by_count":0,"area":null},{"id":"2609.14127","title":"Exact Finite Attention Responses From RoPE Derivatives","created":"2026-09-12","cited_by_count":0,"area":null},{"id":"2609.14174","title":"Inherited Heads: Audio language models track speakers with their text backbone's attention, and an attention-mass ranking retrieves a different set","created":"2026-09-12","cited_by_count":0,"area":null},{"id":"2609.35108","title":"A mechanistic study of language model introspection","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2609.35551","title":"BaRe-Mem: Bayesian Reliability Memory for Robust and Adaptive Agent Consultation","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2609.36434","title":"The Safety Operator: Modulating the Expression of Safety Instructions via Spectral Optimization","created":"2026-09-29","cited_by_count":0,"area":null},{"id":"2609.37263","title":"Beyond Attention Imbalance: Mitigating Hallucinations via Spectral Surgery","created":"2026-09-29","cited_by_count":0,"area":"mllm"},{"id":"2609.38866","title":"When Context Changes: Understanding Update Failures in LLMs","created":"2026-09-30","cited_by_count":0,"area":"agents"},{"id":"2610.02664","title":"A GHOST in Long-Horizon Agents: Governance Hazard from Overlooked Safety Constraints across Turns","created":"2026-10-02","cited_by_count":0,"area":"agents"},{"id":"2610.04329","title":"Suppressing Pressure, Amplifying Evidence: Self-Guided Attention Steering to Mitigate Sycophancy and Stubbornness","created":"2026-10-03","cited_by_count":0,"area":null},{"id":"2610.04676","title":"Extracting Persona Subspaces Through Iterative Nullspace Projection For Modulation","created":"2026-10-03","cited_by_count":0,"area":null},{"id":"2610.05564","title":"Lend Me Your Eyes: Instruction-Aware Text Embeddings via Attention Relay","created":"2026-10-04","cited_by_count":0,"area":null},{"id":"2610.13111","title":"What Do Visual-Head Discovery Scores Actually Find? Attention Source, Necessity, and Random-Control Pitfalls in Vision-Language Models","created":"2026-10-09","cited_by_count":0,"area":"mllm"},{"id":"2610.15883","title":"What the Model Cannot See: Covert and Provable Text Invisibility in LLMs","created":"2026-10-11","cited_by_count":0,"area":"safety"},{"id":"2610.17079","title":"Rethinking LLM-Guided Visual Question Answering: Does Deeper Knowledge Integration Really Help?","created":"2026-10-12","cited_by_count":0,"area":"lr"}],"next_cursor":null}