AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Temporal Context Reinstatement Drives Episodic-Like Order Memory in Long-Context Language Models

Mathis Pink, Vy A. Vo, Qinyuan Wu, Jianing Mu, Javier S. Turek, Uri Hasson, Kenneth A. Norman, Sebastian Michelmann, Alexander Huth, Mariya Toneva

ICML 2026 · Large language models › Language, safety and interpretability · approach: Reasoning mechanisms and failure modes · 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
add the time-reinstatement head at retrieval (vs removing it) (ablation)Mistral-7B-v0.2 · SORTaccuracy: 0.60 -> 0.50Improved
add the time-reinstatement head at retrieval (vs removing it) (ablation)Qwen2.5-7B · SORTaccuracy: 0.65 -> 0.60Improved
add head L8H11 at retrieval (vs removing it), 500-word context (ablation)Llama3.1-8B-Instruct · Dorian Gray orderingKendall tau: 0.63 -> 0.54Improved
add head L8H11 at retrieval (vs removing it), 25k-word context (ablation)Llama3.1-8B-Instruct · Dorian Gray orderingKendall tau: 0.31 -> 0.11Improved
add temporal direction (vs projecting it out) in time-reinstatement head (ablation)Llama3.1-8B and 70B · TMORA SORTaccuracy: 10-15%Improved
amplify gain on L8H11 by factor 1.75Llama3.1-8B-Instruct · TMORA SORTaccuracy: up to 9%Improved
shuffle text in blocks of four sentences (ablation)Llama3.1-8B and 70B · TMORA shuffledaccuracy: p >= 0.307 (8B), p >= 0.104 (70B)No clear effect
disable RoPE (ablation)Llama3.1-8B-Instruct · random tokens, temporal info in value cacheKendall tau: cosine similarity 0.69Improved
scale from 8B to 70B model vs humansLlama3.1-70B-Instruct · TMORA 16k-24k words apartaccuracy: p = 0.021Improved
compare Llama3.1-8B with humansLlama3.1-8B-Instruct · TMORAaccuracy: all p >= 0.289No clear effect

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