mlsanityData to 14 Oct 2026

Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia

arXiv 2608.12717 ↗First posted 13 Aug 2026cs.LG cs.CLcited by 0

On the map: Mechanistic interpretability → Interpreting biomedical and science models

Abstract (arXiv)

Mechanistic interpretability of large language models lacks spatially resolved, falsifiable tools for testing whether internal components are specialized for distinct cognitive operations. We adapt subtraction analysis, the standard framework of human neuroimaging, from biological brains to perturbed transformers, and apply the same logic to both substrates in parallel. Building on the Brain-LLM Unified Model (BLUM), which showed that layer-perturbed LLaVA-1.6-Vicuna-13B error profiles match the lesion patterns of aphasic patients, we develop PRISM (Perturbation-based Regional Interpretability through Subtraction Mapping). PRISM maps the seven clinical Philadelphia Naming Test categories, subtracts error classes pairwise, and treats each perturbation seed as a subject in a group analysis with threshold-free cluster enhancement along the layer axis. We run a structurally matched analysis on 213 chronic post-stroke aphasia patients using correlation-difference lesion-symptom mapping, and replicate both sides on held-out splits. The designs match in subject dimension (seeds, patients), spatial dimension (layers, atlas-parcellated cortex) and thresholding, but the contrast operator differs: a within-subject error-proportion difference for the LLM, a between-subject correlation difference for the cortex. Both substrates recover a robust phonemic-favoring dissociation, a deep layer cluster and a frontal-perisylvian cortical cluster, both replicating; the semantic-favoring direction is a consistently signed but non-significant trend on both. PRISM thus gives a falsifiable, spatially resolved test of functional-specialization claims in transformer language models. A confirmatory ROI-level intervention (PRISM Stage 3) licensing the strongest causal-mechanism claim is left to subsequent work.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Interpreting biomedical and science models, one of 13 directions in Mechanistic interpretability. Builds on 3 of its 22 references that are in the map. First posted on arXiv 13 Aug 2026.

Builds on

3 of 22 references.

References · 3

  1. 2019BERT Rediscovers the Classical NLP Pipelinecited by 1,059
  2. 2026Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnetcited by 191
  3. 2026Stroke Lesions as a Rosetta Stone for Language Model Interpretabilitycited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Recovering Lesion Parameters from Aphasic Picture Naming Error Profiles in Large Language Models
  2. 2026Lesioned Multimodal Language Models Reproduce Aphasic Picture-Naming Patterns
  3. 2026Component-Level Lesioning of Language Models Reveals Clinically Aligned Aphasia Phenotypes
  4. 2026A Monosemantic Attribution Framework for Stable Interpretability in Clinical Neuroscience Transformer-Based Language Models
  5. 2026Sparse Autoencoders Map Brain-LLM Alignment onto Cortical Semantic Topography
  6. 2026Cognitive Expert Language Models Better Align with the Corresponding Brain Systems
  7. 2026Computational Lesions in Multilingual Language Models Separate Shared and Language-specific Brain Alignment
  8. 2026Do Language Models Align with Brains? Prediction Scores Are Not Enough