AI Paper Atlas

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

Is In-Context Learning Learning?

Adrian de Wynter

ICLR 2026 · Large language models › Language, safety and interpretability · approach: In-context learning analysis

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
Increase the number of in-context exemplarsICL (all prompts) · Nine formal tasksaccuracy: slope 8.3 (modus ponens) to 3.3 (CoT)Improved
Increase shots until models convergeICL · Nine formal tasksaccuracy gap between models: -2.6 per shot slope of model gapImproved
Replace natural-language system prompt with random words (word salad) (ablation)ICL · Nine formal tasksaccuracyNo clear effect
Randomise the exemplars and the prompt (Salad-of-Thought) (ablation)CoT · Nine formal tasksaccuracy: average 23 +/- 4Hurt
Shuffle position of the same exemplars (ablation)ICL · Nine formal tasksaccuracyNo clear effect
Fully randomise exemplars drawn i.i.d. from the training set (ablation)GPT-4o · Nine formal tasksaccuracyHurt
Use only negative labels in exemplars (imbalanced) (ablation)GPT-4o · Eight formal tasksaccuracyNo clear effect
Use chain-of-thought promptingICL · Distribution shift, delta to 0.85accuracy: OOD slope -1.4 versus -0.4 for modus ponensMixed
Increase distributional shift between exemplars and testICL · Nine formal tasksaccuracy: average slope -0.6 +/- 0.3Hurt
Use classical baselines instead of ICLkNN, DT, MLP · Nine formal tasksaccuracyMixed
Factor out parsing (compliance) errors (ablation)ICL · Nine formal tasksperceived accuracyMixed
Add chain-of-thought prompt to few-shotCoT · Nine formal tasksaccuracy: best prompt in four tasks; worst in Pattern Matching and HamiltonianMixed

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