AI Paper Atlas

Research map / Probabilistic and scientific modeling / Dynamical systems

Towards Spectroscopy: Susceptibility Clusters in Language Models

Andrew Gordon, Garrett Baker, George Wang, William Snell, Stan van Wingerden, Daniel Murfet

ICML 2026 · Probabilistic and scientific modeling › Dynamical systems · approach: Physics-informed dynamics learning · 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
evaluate Pythia-14M clusters with susceptibilities from larger models (31M to 1.4B) instead of random vertex sets (ablation)Pythia-14M susceptibility clusters · conductance in larger Pythia modelsconductance: random sets approx 1 -> clusters significantly below 1Improved
cluster larger models directly instead of Pythia-14M (ablation)conductance clustering · number of clustersclusters found: 510 -> 241--358Hurt
compare clusters with SAE featuresPythia-14M clusters · Pythia-70M SAE feature matchmatch rateImproved
assign tokens by function: same y token in different rolessusceptibility vectors · context sensitivity (e.g. * as operator vs markdown)cluster separationImproved

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