{"total":73,"rows":[{"id":"2410.10781","title":"When Attention Sink Emerges in Language Models: An Empirical View","created":"2024-10-14","cited_by_count":208,"area":"inference"},{"id":"2407.00079","title":"Mooncake: A KVCache-centric Disaggregated Architecture for LLM Serving","created":"2024-06-24","cited_by_count":196,"area":"inference"},{"id":"2503.02175","title":"DivPrune: Diversity-based Visual Token Pruning for Large Multimodal Models","created":"2025-03-04","cited_by_count":181,"area":"mllm"},{"id":"2410.13835","title":"Active-Dormant Attention Heads: Mechanistically Demystifying Extreme-Token Phenomena in LLMs","created":"2024-10-17","cited_by_count":56,"area":null},{"id":"2410.06270","title":"Mixture Compressor for Mixture-of-Experts LLMs Gains More","created":"2024-10-08","cited_by_count":41,"area":"moe"},{"id":"2504.02441","title":"Cognitive Memory in Large Language Models","created":"2025-04-03","cited_by_count":29,"area":null},{"id":"2504.17768","title":"The Sparse Frontier: Sparse Attention Trade-offs in Transformer LLMs","created":"2025-04-24","cited_by_count":29,"area":null},{"id":"2410.17897","title":"Value Residual Learning","created":"2024-10-23","cited_by_count":27,"area":"inference"},{"id":"2601.14004","title":"Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models","created":"2026-01-20","cited_by_count":27,"area":"mechinterp"},{"id":"2604.04921","title":"TriAttention: Efficient Long Reasoning with Trigonometric KV Compression","created":"2026-04-06","cited_by_count":21,"area":"lr"},{"id":"2506.04213","title":"FullDiT2: Efficient In-Context Conditioning for Video Diffusion Transformers","created":"2025-06-04","cited_by_count":18,"area":"diffusion"},{"id":"2507.08143","title":"Compactor: Calibrated Query-Agnostic KV Cache Compression with Approximate Leverage Scores","created":"2025-07-10","cited_by_count":17,"area":"inference"},{"id":"2502.00919","title":"Attention Sinks: A 'Catch, Tag, Release' Mechanism for Embeddings","created":"2025-02-02","cited_by_count":16,"area":null},{"id":"2502.01941","title":"Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression","created":"2025-02-04","cited_by_count":16,"area":"lr"},{"id":"2509.09754","title":"LAVa: Layer-wise KV Cache Eviction with Dynamic Budget Allocation","created":"2025-09-11","cited_by_count":16,"area":"inference"},{"id":"2411.13476","title":"When Precision Meets Position: BFloat16 Breaks Down RoPE in Long-Context Training","created":"2024-11-20","cited_by_count":15,"area":null},{"id":"2505.20334","title":"Lookahead Q-Cache: Achieving More Consistent KV Cache Eviction via Pseudo Query","created":"2025-05-24","cited_by_count":14,"area":"inference"},{"id":"2407.15176","title":"ReAttention: Training-Free Infinite Context with Finite Attention Scope","created":"2024-07-21","cited_by_count":13,"area":null},{"id":"2505.10118","title":"Why 1 + 1 < 1 in Visual Token Pruning: Beyond Naive Integration via Multi-Objective Balanced Covering","created":"2025-05-15","cited_by_count":11,"area":"inference"},{"id":"2510.07651","title":"OBCache: Optimal Brain KV Cache Pruning for Efficient Long-Context LLM Inference","created":"2025-10-09","cited_by_count":10,"area":"inference"},{"id":"2603.10899","title":"LookaheadKV: Fast and Accurate KV Cache Eviction by Glimpsing into the Future without Generation","created":"2026-03-11","cited_by_count":10,"area":"inference"},{"id":"2512.17914","title":"Q-KVComm: Efficient Multi-Agent Communication Via Adaptive KV Cache Compression","created":"2025-11-27","cited_by_count":9,"area":"agents"},{"id":"2605.08840","title":"ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing","created":"2026-05-09","cited_by_count":9,"area":"inference"},{"id":"2504.03889","title":"Identifying and Evaluating Inactive Heads in Pretrained LLMs","created":"2025-04-04","cited_by_count":8,"area":null},{"id":"2505.16175","title":"QuickVideo: Real-Time Long Video Understanding with System Algorithm Co-Design","created":"2025-05-22","cited_by_count":8,"area":null},{"id":"2606.03928","title":"Value-Aware Stochastic KV Cache Eviction for Reasoning Models","created":"2026-06-02","cited_by_count":8,"area":"lr"},{"id":"2412.16187","title":"HashEvict: A Pre-Attention KV Cache Eviction Strategy using Locality-Sensitive Hashing","created":"2024-12-13","cited_by_count":6,"area":"inference"},{"id":"2505.14884","title":"Polar Sparsity: High Throughput Batched LLM Inferencing with Scalable Contextual Sparsity","created":"2025-05-20","cited_by_count":6,"area":null},{"id":"2509.00388","title":"GraphKV: Breaking the Static Selection Paradigm with Graph-Based KV Cache Eviction","created":"2025-08-30","cited_by_count":6,"area":"inference"},{"id":"2509.24832","title":"SemShareKV: Efficient KVCache Sharing for Semantically Similar Prompts via Token-Level LSH Matching","created":"2025-09-29","cited_by_count":6,"area":"inference"},{"id":"2510.17021","title":"Forgetting to Forget: Attention Sink as A Gateway for Backdooring LLM Unlearning","created":"2025-10-19","cited_by_count":6,"area":"safety"},{"id":"2601.21531","title":"On the Adversarial Robustness of Large Vision-Language Models under Visual Token Compression","created":"2026-01-29","cited_by_count":6,"area":"mllm"},{"id":"2603.11504","title":"LongFlow: Efficient KV Cache Compression for Reasoning Models","created":"2026-03-12","cited_by_count":6,"area":"lr"},{"id":"2408.01890","title":"Cross-layer Attention Sharing for Pre-trained Large Language Models","created":"2024-08-04","cited_by_count":5,"area":"inference"},{"id":"2605.08234","title":"When Does Value-Aware KV Eviction Help? A Fixed-Contract Diagnostic for Non-Monotone Cache Compression","created":"2026-05-07","cited_by_count":5,"area":"inference"},{"id":"2605.19660","title":"OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond","created":"2026-05-19","cited_by_count":5,"area":"inference"},{"id":"2503.02812","title":"Q-Filters: Leveraging QK Geometry for Efficient KV Cache Compression","created":"2025-03-04","cited_by_count":4,"area":"inference"},{"id":"2603.08727","title":"ARKV: Adaptive and Resource-Efficient KV Cache Management under Limited Memory Budget for Long-Context Inference in LLMs","created":"2026-02-19","cited_by_count":4,"area":"agents"},{"id":"2607.08057","title":"Towards Efficient Large Language Model Serving: A Survey on System-Aware KV Cache Optimization","created":"2026-07-09","cited_by_count":4,"area":"inference"},{"id":"2412.05896","title":"XKV: Personalized KV Cache Memory Reduction for Long-Context LLM Inference","created":"2024-12-08","cited_by_count":3,"area":"inference"},{"id":"2502.19078","title":"Sparse Brains are Also Adaptive Brains: Cognitive-Load-Aware Dynamic Activation for LLMs","created":"2025-02-26","cited_by_count":3,"area":null},{"id":"2507.22798","title":"Quantifying surprise in clinical care: Detecting highly informative events in electronic health records with foundation models","created":"2025-07-30","cited_by_count":3,"area":null},{"id":"2602.20205","title":"OTPrune: Distribution-Aligned Visual Token Pruning via Optimal Transport","created":"2026-02-22","cited_by_count":3,"area":"lr"},{"id":"2604.24647","title":"DepthKV: Layer-Dependent KV Cache Pruning for Long-Context LLM Inference","created":"2026-04-27","cited_by_count":3,"area":"inference"},{"id":"2410.12850","title":"RecurFormer: Not All Transformer Heads Need Self-Attention","created":"2024-10-10","cited_by_count":2,"area":null},{"id":"2510.04547","title":"Activation Quantization of Vision Encoders Needs Prefixing Registers","created":"2025-10-06","cited_by_count":2,"area":"mllm"},{"id":"2511.14182","title":"WebRec: Enhancing LLM-based Recommendations with Attention-guided RAG from Web","created":"2025-11-18","cited_by_count":2,"area":"rag"},{"id":"2602.08686","title":"CompilerKV: Risk-Adaptive KV Compression via Offline Experience Compilation","created":"2026-02-09","cited_by_count":2,"area":null},{"id":"2603.06248","title":"Gradient Flow Polarizes Softmax Outputs towards Low-Entropy Solutions","created":"2026-03-06","cited_by_count":2,"area":null},{"id":"2508.12590","title":"Energy-Efficient Wireless LLM Inference via Uncertainty and Importance-Aware Speculative Decoding","created":"2025-08-18","cited_by_count":1,"area":"inference"},{"id":"2602.00247","title":"CAPA: Contribution-Aware Pruning and FFN Approximation for Efficient Large Vision-Language Models","created":"2026-01-30","cited_by_count":1,"area":"mllm"},{"id":"2602.01984","title":"Enhancing Multi-Image Understanding through Delimiter Token Scaling","created":"2026-02-02","cited_by_count":1,"area":"mllm"},{"id":"2604.01178","title":"Screening Is Enough","created":"2026-04-01","cited_by_count":1,"area":null},{"id":"2604.18103","title":"Stability Implies Redundancy: Delta Attention Selective Halting for Efficient Long-Context Prefilling","created":"2026-04-20","cited_by_count":1,"area":"mllm"},{"id":"2605.09877","title":"Key-Value Means: Transformers with Expandable Block-Recurrent Compressed Memory","created":"2026-05-11","cited_by_count":1,"area":null},{"id":"2607.24555","title":"LOCKS: Page-Local Compact Key Summaries for Efficient Long-Context Decoding","created":"2026-07-27","cited_by_count":1,"area":null},{"id":"2608.03276","title":"TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning","created":"2026-08-04","cited_by_count":1,"area":"inference"},{"id":"2608.14191","title":"KV Cache Compression Through the Lens of Transform Coding","created":"2026-08-14","cited_by_count":1,"area":"inference"},{"id":"2608.25230","title":"Trust the Mass: Forced Weights in KV-Cache Eviction","created":"2026-08-25","cited_by_count":1,"area":"inference"},{"id":"2609.36835","title":"ARC-KV: Amortizing Anchor Search for Reconstruction-Based KV Cache Compaction","created":"2026-09-29","cited_by_count":1,"area":"inference"},{"id":"2502.13176","title":"BaKlaVa -- Budgeted Allocation of KV cache for Long-context Inference","created":"2025-02-18","cited_by_count":0,"area":"inference"},{"id":"2510.11789","title":"Minimax Rates for Learning Pairwise Interactions in Attention-Style Models","created":"2025-10-13","cited_by_count":0,"area":null},{"id":"2511.16943","title":"RASTP: Representation-Aware Semantic Token Pruning for Generative Recommendation with Semantic Identifiers","created":"2025-11-21","cited_by_count":0,"area":"inference"},{"id":"2604.00757","title":"IWP: Token Pruning as Implicit Weight Pruning in Large Vision Language Models","created":"2026-04-01","cited_by_count":0,"area":"mllm"},{"id":"2606.00024","title":"ART: Attention Run-time Termination for Efficient Large Language Model Decoding","created":"2026-04-15","cited_by_count":0,"area":null},{"id":"2606.07604","title":"Contribution Weights: A Geometrical Analysis of Self-Attention Transformers","created":"2026-05-29","cited_by_count":0,"area":null},{"id":"2607.16213","title":"SelKV: Selective KV Cache Merging with Per-Token Merge-or-Drop and Attention Compensation","created":"2026-05-19","cited_by_count":0,"area":"inference"},{"id":"2608.03112","title":"Adaptive Two-Stage Visual Token Pruning for Efficient Inference in Video-Language Models","created":"2026-08-04","cited_by_count":0,"area":"mllm"},{"id":"2608.21541","title":"Beyond Sparse Weights: When Is Attention Compressible?","created":"2026-08-21","cited_by_count":0,"area":"inference"},{"id":"2608.27128","title":"Small Frequency Corrections Can Change What Survives KV Cache Compression","created":"2026-08-27","cited_by_count":0,"area":"inference"},{"id":"2609.26052","title":"Optimizing Denoising Trajectories in dLLMs: A Lightweight Evolutionary Heuristic Approach","created":"2026-07-15","cited_by_count":0,"area":null},{"id":"2609.35593","title":"ReSS: Residual-Restoring Sparse Attention for 3D Vision Transformers","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2610.13337","title":"Exploring Attention Sinks in Text-to-Image Transformers from Diffusion and Autoregressive Models","created":"2026-10-09","cited_by_count":0,"area":null}],"next_cursor":null}