AI Paper Atlas

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

Controlling the Risk of Corrupted Contexts for Language Models via Early-Exiting

Andrea Wynn, Metod Jazbec, Charith Peris, Rinat Khaziev, Anqi Liu, Daniel Khashabi, Eric Nalisnick

ICML 2026 · Large language models › Language, safety and interpretability · approach: LLM pruning and scaling data · 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
early-exit at an intermediate layer instead of using the full model under harmful contextLlama-style LLMs with ICL · ICL classification tasks with incorrect demonstrationsaccuracyImproved
early exit with a suitable threshold also on helpful contextearly-exit LLM · ICL tasks with correct demonstrationsaccuracyMixed
replace loss clipping with domain-preserving risk transformation in LTT (ablation)risk-controlled early exit · 8 ICL tasks (epsilon 0.05)layers evaluated: 53% fewer layers on averageImproved
replace loss clipping with domain-preserving risk transformation in LTT (ablation)risk-controlled early exit · SQuAD v2.0layers evaluated: up to 81.8% fewer layersImproved
replace loss clipping with risk transformation (ablation)risk-controlled early exit · ICL tasks and SQuAD v2.0risk control validityNo clear effect
add zero-shot fallback when no exit is confidentearly-exit model · SQuAD v2.0fallback rate on misleading vs answerable questions: 3.0% vs 0.9%Improved
fall back to zero-shot on unsafe context rather than full-model prediction (as CALM does)early-exit model · SQuAD v2.0 unanswerablequestion-level correctness: zero-shot better on 71% of questionsImproved
restrict early exit to the second half of layers (ablation)early-exit design · all tasks and modelssafety/accuracyImproved

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