ARC-KV: Amortizing Anchor Search for Reconstruction-Based KV Cache Compaction
cs.AI
Submitted: 2026-09-29
Updated: 2026-09-29
Terminology
Sources
- PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
- Compactor: Calibrated Query-Agnostic KV Cache Compression with Approximate Leverage Scores
- KV-Distill: Nearly Lossless Learnable Context Compression for LLMs
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- Expected Attention: KV Cache Compression by Estimating Attention from Future Queries Distribution
- CAOTE: KV Cache Selection for LLMs via Attention Output Error-Based Token Eviction
- The Llama 3 Herd of Models
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- RULER: What's the Real Context Size of Your Long-Context Language Models?
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- Learning to Evict from Key-Value Cache
- Still: Amortized KV Cache Compaction in a Single Forward Pass
- KV-Compress: Paged KV-Cache Compression with Variable Compression Rates per Attention Head
- Gemma 3 Technical Report
- Qwen3 Technical Report
- IndexMem: Learned KV-Cache Eviction with Latent Memory for Long-Context LLM Inference
- In-context KV-Cache Eviction for LLMs via Attention-Gate
- LKV: End-to-End Learning of Head-wise Budgets and Token Selection for LLM KV Cache Eviction
- Fast KV Compaction via Attention Matching
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems