AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

The First Drop of Ink: Nonlinear Impact of Misleading Information in Long-Context Reasoning

Muhan Gao, Zih-Ching Chen, Kuan-Hao Huang

ICML 2026 · Large language models › Retrieval and knowledge · approach: Retrieval-augmented reasoning

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 the first 10% of hard distractors to a fixed-length 128K context (ablation)Llama-3.1-8B-Instruct · nq_easy 128Kaccuracy (%): 87.0 -> 72.5Hurt
raise hard distractor share from 10% to 100% in 128K context (ablation)Llama-3.1-8B-Instruct · nq_easy 128Kaccuracy (%): 72.5 -> 62.0Hurt
add the first 10% of hard distractors to a fixed-length 128K context (ablation)Qwen2.5-7B-Instruct · nq_easy 128Kdrop ratio: drop ratio 0.58 vs 0.1 for linearHurt
lower softmax temperature at inference (tau < 1) to sharpen attention on the target (ablation)Llama-3.1-8B-Instruct · nq_easyaccuracyHurt
filter hard distractors instead of weaker (random) distractors, from 131K down to 47K tokens (ablation)Llama-3.1-8B-Instruct and Qwen2.5-7B-Instruct · multi-document QA, four datasetsaccuracyNo clear effect
filter hard distractors until their share is near zero (3%) at 27K tokens, versus filtering random (ablation)Llama-3.1-8B-Instruct and Qwen2.5-7B-Instruct · multi-document QA, four datasetsaccuracyImproved
reduce context length from 131K to 27K while keeping hard ratio fixed at 20%, 50% or 80% (ablation)Llama-3.1-8B-Instruct and Qwen2.5-7B-Instruct · multi-document QA, four datasetsaccuracyImproved
vary the fixed hard distractor ratio among 20%, 50%, 80% during proportional context reduction (ablation)Llama-3.1-8B-Instruct and Qwen2.5-7B-Instruct · multi-document QA, four datasetsaccuracyNo clear effect
use LLM judge (gpt-4o-mini) instead of string containment matchingevaluation of QA accuracy · HotpotQAfalse negatives: 36 of 200 false negatives under string matching vs 17 in total over 100 checked samples per dataset (4.25%)Improved

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