AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing

Yongqi An, Chang Lu, Kuan Zhu, Tao Yu, Chaoyang Zhao, Hong Wu, Ming Tang, Jinqiao Wang

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: KV cache eviction and compression · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add layer-wise output reconstruction indicator (ablation)ReST-KV · LongBench (128L)avg score: 33.95 -> 35.86Improved
add EMA temporal smoothing (ablation)ReST-KV · LongBench (128L)avg score: 34.02 -> 35.86Improved
add adaptive window spatial smoothing (ablation)ReST-KV · LongBench (128L)avg score: 33.50 -> 35.86Improved
use full ReST-KV instead of attention-weight top-k (ablation)KV eviction · LongBench (128L)avg score: 32.98 -> 35.86Improved
combine ReST-KV with PyramidKV layer-wise budget allocation (ablation)ReST-KV · LongBench (Llama2-7B-Chat, avg over budgets)avg score: 30.48 -> 30.93Improved
combine ReST-KV with AdaKV head-wise budget allocation (ablation)ReST-KV · LongBench (Llama2-7B-Chat, avg over budgets)avg score: 30.48 -> 31.20Improved
use ReST-KV instead of SnapKV at tight budgetKV eviction · LongBench (Llama-3.1-8B, 64L)avg score: 31.48 -> 33.54Improved
use ReST-KV instead of SnapKV at large budgetKV eviction · LongBench (Llama2-7B-Chat, 1024L, with Pyramid/Ada table)avg score: 32.99 -> 32.97No clear effect
use ReST-KV instead of SnapKV on RULERKV eviction · RULER (avg 4K-128K)avg accuracy: 67.11 -> 82.27Improved
ReST-KV prefill overhead vs SnapKVReST-KV · Prefill TTFT 128kTTFT (ms): 28,510 -> 29,082Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy