Sparse Attention Is Matrix Approximation, Not Choosing from a Bag of Values
cs.CL
Submitted: 2026-10-07
Updated: 2026-10-07
Terminology
Sources
- Beyond Efficiency: A Systematic Survey of Resource-Efficient Large Language Models
- Longformer: The Long-Document Transformer
- Generating Long Sequences with Sparse Transformers
- CAOTE: KV Cache Selection for LLMs via Attention Output Error-Based Token Eviction
- Mamba: Linear-Time Sequence Modeling with Selective State Spaces
- RULER: What's the Real Context Size of Your Long-Context Language Models?
- KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache
- Retentive Network: A Successor to Transformer for Large Language Models
- Gemini: A Family of Highly Capable Multimodal Models
- Tactic: Adaptive Sparse Attention with Clustering and Distribution Fitting for Long-Context LLMs
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering