AI Paper Atlas

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

How Does the Pretraining Distribution Shape In-Context Learning? A Fundamental Trade-Off

Waïss Azizian, Ali Hasan

ICML 2026 · Large language models › Language, safety and interpretability · approach: In-context learning analysis · 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 heavier-tailed task prior (lower Student-t degrees of freedom) in pretraining (ablation)transformer ICL, linear regression, N=infinity · linear regression under task shiftlog MSEMixed
use heavier-tailed prior with small number of pretraining tasks (N=100) (ablation)transformer ICL, linear regression · linear regression under task shift, finite Nlog MSEHurt
use heavier-tailed prior (lower nu or beta) in pretraining (ablation)transformer ICL, Ornstein-Uhlenbeck next-step prediction · OU process under task shiftMSEImproved
stronger temporal dependence (smaller Volterra kernel exponent) (ablation)transformer ICL, Volterra process · Volterra equations, N from 5000 to 500MSE: MSE at shift 0 multiplied by 3 for exponent 1.0; barely changes for exponent 2.0Hurt
importance-reweight pretraining tasks toward uniform over observed range (ablation)Student-t pretraining prior, linear regression and OU · task shift evaluationMSE: small improvement under large shiftsImproved
importance-reweight pretraining tasks (ablation)generalized normal prior, OU process · OU process task shiftMSENo 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