Metraipolis › archive › dc_7d95e42a7cd3

Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia — arXiv:2608.12717

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.12717
License (as archived)
CC0-1.0
Archived
2026-10-19T04:24:40.723Z
SHA-256 of the archived text
e5f9cc9ea4081fe4def46dac2817c793a11997007f91ae1ffdee9c028d004cd3
Stable URL of this page
https://metraipolis.com/documento/dc_7d95e42a7cd3

How to cite

Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia — arXiv:2608.12717 — arXiv (en), https://arxiv.org/abs/2608.12717 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7d95e42a7cd3 (Metraipolis archive dc_7d95e42a7cd3, sha256 e5f9cc9ea4081fe4def46dac2817c793a11997007f91ae1ffdee9c028d004cd3)
BibTeX
@misc{dc7d95e42a7cd3,
  title = {Perturbation-based Regional Interpretability through Subtraction Mapping (PRISM): naming-error dissociations in language models and post-stroke aphasia — arXiv:2608.12717},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.12717},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7d95e42a7cd3; sha256 e5f9cc9ea4081fe4def46dac2817c793a11997007f91ae1ffdee9c028d004cd3},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12717. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.CL).
Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.12717 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T04:17:42.408Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).