AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

InfoFlow KV: Information-Flow-Aware KV Recomputation for Long Context

Xin Teng, Canyu Zhang, Shaoyi Zheng, Danyang Zhuo, Tianyi Zhou, Shenji Wan

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Multimodal retrieval-augmented modeling · 0 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
use GLOBAL RoPE geometry instead of HL--TP for token selection (ablation)Our · 2WikiMQA (passage split)F1: 0.4722 -> 0.5019Improved
use GLOBAL RoPE geometry instead of HL--HP (ablation)Our · 2WikiMQA (passage split)F1: 0.4455 -> 0.5019Improved
use GLOBAL RoPE geometry instead of TL--TP (ablation)Our · NarrativeQA (passage split)F1: 0.1923 -> 0.2288Improved
use GLOBAL RoPE geometry instead of HL--HP (ablation)Our · HotpotQA (passage split)F1: 0.5529 -> 0.5954Improved
place important chunks closer to query (ascending) instead of descending (ablation)Our + Reorder · 2WikiMQA (passage split)F1: 0.4607 -> 0.5058Improved
place important chunks closer to query (ascending) instead of descending (ablation)Our + Reorder · MuSiQue (passage split)F1: 0.2875 -> 0.3285Improved
place important chunks closer to query (ascending) instead of descending (ablation)Our + Reorder · HotpotQA (passage split)F1: 0.6054 -> 0.5972Hurt
add chunk reordering to OurOur, Qwen3-14B · 2WikiMQA (passage split)F1: 0.5019 -> 0.5058No clear effect
add chunk reordering to OurOur, Qwen3-14B · MuSiQue (passage split)F1: 0.3386 -> 0.3285Hurt
add chunk reordering to OurOur, Qwen3-14B · 2WikiMQA (fixed chunk 2048)F1: 0.5089 -> 0.4773Hurt
use norm-based selection instead of CacheBlendQwen3-14B recomputation · 2WikiMQA (passage split)F1: 0.4330 -> 0.5019Improved
use token-level selection instead of BM25 passage-level recomputeQwen3-14B recomputation · 2WikiMQAF1: 0.1904 -> 0.5019Improved

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