Metraipolis › archive › dc_52898daa70fb

Concentrated Causal Influence, Distributed Diagnostic Information: Cognitive Impairment Decisions in Large Audio-Language Models — arXiv:2610.32061

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.32061
License (as archived): CC0-1.0
Archived: 2026-10-26T19:53:48.224Z
SHA-256 of the archived text: c69aa15f6aee4965cab3d181a991fb2d86b52a97c0f29c1d6ed58fda73972551
Stable URL of this page: https://metraipolis.com/documento/dc_52898daa70fb

How to cite

Concentrated Causal Influence, Distributed Diagnostic Information: Cognitive Impairment Decisions in Large Audio-Language Models — arXiv:2610.32061 — arXiv (en), https://arxiv.org/abs/2610.32061 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_52898daa70fb (Metraipolis archive dc_52898daa70fb, sha256 c69aa15f6aee4965cab3d181a991fb2d86b52a97c0f29c1d6ed58fda73972551)

BibTeX

@misc{dc52898daa70fb,
  title = {Concentrated Causal Influence, Distributed Diagnostic Information: Cognitive Impairment Decisions in Large Audio-Language Models — arXiv:2610.32061},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.32061},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_52898daa70fb; sha256 c69aa15f6aee4965cab3d181a991fb2d86b52a97c0f29c1d6ed58fda73972551},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.32061. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large audio-language models (LALMs) can assess cognitive impairment (CI) from speech, yet their answers reveal neither which internal components control the decision nor how much CI information is accessible in their representations. We investigate both questions through CI-contrast neurons (CICNs), selected by class-contrastive activation frequency in decoder MLPs. Across eight open LALMs and two corpora (Pitt and PROCESS-2), we causally test CICNs against size- and layer-matched random masks and activity-matched controls. In three of the five models that meet our eligibility criteria, ablation of small CICN sets causes large, class-directed prediction shifts; speech-identified masks also control transcript-only decisions, and three of four primary Pitt masks retain specificity on PROCESS-2. Position-restricted interventions locate the influence before the final answer position, most strongly after the speech or transcript evidence under letter-labeled options, largely as a decision-direction shift. Beyond causal localization, supervised linear readouts of frozen CICN activations achieve 0.92--0.97 AUC within Pitt, versus 0.67--0.84 for the models' direct answers. Matched random and all-neuron readouts reach comparable performance; top-ranked CICNs show only an exploratory advantage at very small unit budgets. These results distinguish concentrated causal control over output decisions from broadly distributed diagnostic information for CI detection in LALMs.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Audio and Speech Processing (eess.AS) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.32061 — 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 2026-10-26T19:45:48.386Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).