AI Paper Atlas

Research map / Label-efficient and robust learning / Test-time adaptation

Learning to Extrapolate to New Tasks: A Relational Approach to Task Extrapolation

Adam Ousherovitch, Yixin Wang

ICML 2026 · Label-efficient and robust learning › Test-time adaptation · approach: Extrapolation and single-sample adaptation · 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
decompose target into anchor plus transformation (RTE) instead of inductive predictionMLP · Parameter extrapolation, CubicMSE: 2.86 -> 1.53Improved
decompose target into anchor plus transformation (RTE)MLP · Parameter extrapolation, Sin TrendMSE: 0.28 -> 0.051Improved
decompose target into anchor plus transformation (RTE)MLP · Parameter extrapolation, Tri TrendMSE: 0.46 -> 0.048Improved
decompose target into anchor plus transformation (RTE)MLP · Parameter extrapolation, ExpMSE: 1.25 -> 0.80Improved
use inferred anchor and transformation instead of ground-truth ones (ablation)RTE · Parameter extrapolation, Tri TrendMSE: 0.048 vs oracle 0.094No clear effect
add RTE to length extrapolationMLP · Degree 9 polynomialMSE: 0.5752 -> 0.3712Improved
use ground-truth anchor instead of inferred one (ablation)RTE · Degree 9 polynomialMSE: 0.3712 -> 0.0497Improved
add RTE to compositional extrapolationMLP · Unseen compositionsaggregate MSE: 0.3891 -> 0.2867Improved
add relational fine-tuning with anchor and shift promptsLoRA-tuned Qwen/Mistral · Sparse Parity |S|=6accuracy: 52.86% -> 66.07%Improved
add RTE relational promptingLoRA-tuned Qwen/Mistral · CodeIOexact match accuracy: 19.8% -> 45.3%Improved
compare RTE to CoT with majority vote (N=16) (ablation)LLM · CodeIOexact match accuracy: 30.2% -> 45.3%Improved
provide ground-truth decomposition instead of inferred one (ablation)RTE · CodeIOexact match accuracy: 45.3% -> 76.0%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