AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Do LLMs Forget What They Should? Evaluating In-Context Forgetting in Large Language Models

Yuli Qian, Zechuan Yang, Wenbiao Ding, Hongzhi Li, Yutao Xie

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in LLMs

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 explicit forgetting reminder to prompt (ablation)GPT-O3Mini · Instructional ForgettingSFRR: 62.88 -> 76.30Improved
Add explicit forgetting reminder to prompt (ablation)GPT-4o · Instructional ForgettingSFRR: 52.52 -> 71.76Improved
Add explicit forgetting reminder to prompt (ablation)GPT-5 · Instructional ForgettingSFRR: 58.80 -> 74.62Improved
Add explicit forgetting reminder to prompt (ablation)GPT-4o · Dynamic PreferenceSFRR: 50.07 -> 66.56Improved
Add explicit forgetting reminder to prompt (ablation)GPT-5 · Dynamic PreferenceSFRR: 79.12 -> 76.64Hurt
Add explicit forgetting reminder to prompt (ablation)GPT-O3Mini · Instructional ForgettingNoForget Accuracy: 96.48 -> 93.84Hurt
Add explicit forgetting reminder to prompt (ablation)GPT-4o · Dynamic PreferenceNoForget Accuracy: 86.10 -> 87.43Improved
Lengthen context from 0.5k to 30k tokens (ablation)GPT-4o · Instructional ForgettingNoForget Accuracy: 96.53% -> 77.89%Hurt
Lengthen context from 0.5k to 30k tokens (ablation)GPT-4o · Dynamic PreferenceNoForget Accuracy: 90.36% -> 59.42%Hurt
Lengthen context from 0.5k to 30k tokens (ablation)GPT-4o · Dynamic PreferenceSFRR: 52.46% -> 22.87%Hurt
Lengthen context from 0.5k to 30k tokens (ablation)GPT-4o · Subtask RevisionSFRR: 71.76% -> 81.50%Improved
Scale model size (Mistral-7B to GPT-5) on NoForget AccuracyLLMs · ICF-Bench averageNoForget Accuracy: 60.18% -> 88.75%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