Metraipolis › archive › dc_0de93e40da29
Which Attention Heads are like the Human Head? Not the Ones that Compute — arXiv:2609.37991
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37991
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:24:22.407Z
- SHA-256 of the archived text
06f828b8d144a3c8c1965e48583091e90e750ed9db288df0eab8ad8619c973e4- Stable URL of this page
https://metraipolis.com/documento/dc_0de93e40da29
How to cite
Which Attention Heads are like the Human Head? Not the Ones that Compute — arXiv:2609.37991 — arXiv (en), https://arxiv.org/abs/2609.37991 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0de93e40da29 (Metraipolis archive dc_0de93e40da29, sha256 06f828b8d144a3c8c1965e48583091e90e750ed9db288df0eab8ad8619c973e4)
BibTeX
@misc{dc0de93e40da29,
title = {Which Attention Heads are like the Human Head? Not the Ones that Compute — arXiv:2609.37991},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37991},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0de93e40da29; sha256 06f828b8d144a3c8c1965e48583091e90e750ed9db288df0eab8ad8619c973e4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37991. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Brain-AI alignment is often interpreted as a sign that model and brain perform similar computations. Whether the aligned units are causally involved in model computation is rarely checked. On an abstract pattern-completion task (AAABAAA $\rightarrow$ B), we compare LLM attention-head representations with human EEG and test how ablating those heads affects task performance. Alignment and causation dissociate: brain-aligned heads contribute to performance, but their removal is substantially less disruptive than removal of heads selected via attribution patching. We compare two head sets that prior interpretability work defines without reference to the brain: concept vectors (CVs), which represent abstract patterns across formats, and function vectors (FVs), selected for their contribution to correct-answer prediction. Brain alignment shows little association with FV scores, while its association with CV scores varies across models. Among brain-aligned heads, we find recurring attention profiles: one emphasizes distinctive elements (novelty heads), the other repeating elements (repetition heads). The novelty family tracks salience and attends to the same elements that humans look at, yet its removal is less damaging than random ablation on average. Repetition heads contribute modestly to performance and are associated with abstract-pattern representation (CVs). Across 17 models spanning 3B-72B parameters, FV-ranked removal is substantially more disruptive than brain-ranked removal. Brain alignment thus captures how the model reads the stimulus, and only faintly captures how it represents the pattern and solves the task. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Neurons and Cognition (q-bio.NC). Pagina del lavoro: https://arxiv.org/abs/2609.37991 — 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 2026-10-07T15:54:53.678Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).