AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Provably Tracking Equivalent Mechanistic Interpretations Across Neural Networks

Alan Sun, Mariya Toneva

ICLR 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Theory of neural network sensitivity

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
compute congruity between models with same vs different interpretation on n-permutation detectionCongruity · 10-permutation detection, 6 hard-coded transformerscongruity: within-group 0.43 vs across-group 0.01Improved
compare Pythia vs GPT2 models on IOI with CongruityCongruity · IOI, 85M to 2.9B parameterscongruity: across-group 0.13 vs within 0.73 and 0.92Improved
restrict next-token prediction to terminal punctuation or closing brackets/quotationsGPT2 vs POS circuit · C4 next-token predictioncongruity with POS: all-token control 0.48; medium effect sizeImproved
restrict next-token prediction to articles or prepositionsGPT2 vs POS circuit · C4 next-token predictioncongruity with POSNo 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