AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Abstraction Induces the Brain Alignment of Language and Speech Models

Emily Cheng, Aditya R. Vaidya, Richard Antonello

ICML 2026 · Architectures and efficiency › RNNs and normalization · approach: Layer-wise analysis of language models · 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 intrinsic dimension instead of surprisal to explain layerwise encodinglayerwise brain encoding of LLMs and speech models · fMRI, all modelsSpearman rho with encoding performance: -0.53 -> 0.72Improved
use intrinsic dimension instead of surprisal to explain layerwise encodinglayerwise brain encoding · ECoG, all modelsSpearman rho with encoding performance: 0.21 -> 0.43Improved
add I_d to a mixed-effects model with surprisal (ablation)mixed-effects model of encoding performance · fMRIeffect size: beta_surprisal approx 0 vs beta_Id = 0.63Improved
add I_d to a mixed-effects model with surprisal (ablation)mixed-effects model of encoding performance · ECoGeffect size: beta_surprisal = 0.26 vs beta_Id = 0.97Improved
fine-tune WavLM-base-plus layer 9 on fMRI responsesWavLM-base-plus · fMRIencoding performance, semantic content and I_dImproved
replace trained model layers with random Fourier features of matched I_d (ablation)random features · fMRIencoding performance: R approx 0.04 vs R > 0.1Hurt
train Pythia-6.9b longer (more checkpoints)Pythia-6.9b · fMRIcorrelation of I_d with encoding performance: rho = 0.96Improved
train Pythia-6.9b longer (more checkpoints)Pythia-6.9b · ECoGcorrelation of I_d with encoding performance: rho = 0.64Improved
change GRIDE scale k by factor 2 (ablation)I_d estimator · fMRI/ECoGI_d peak layer and correlation: Pythia-6.9b peak layer 13 unchangedNo clear effect
use linear dimension (PR) instead of nonlinear I_d (ablation)dimensionality measure · fMRI, Pythia-6.9bcorrelation with encoding performance: 0.86 (PCA) / -0.05 (PR)Hurt

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