Metraipolis › archive › dc_ef07069c81c4

Signatures of semantic search in the activations of large language models — arXiv:2609.35599

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35599
License (as archived)
CC0-1.0
Archived
2026-10-07T20:58:56.914Z
SHA-256 of the archived text
8d294334ee8e5bf0844b5a78de7bb5fc107e347aa331db557492660cee4f187f
Stable URL of this page
https://metraipolis.com/documento/dc_ef07069c81c4

How to cite

Signatures of semantic search in the activations of large language models — arXiv:2609.35599 — arXiv (en), https://arxiv.org/abs/2609.35599 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ef07069c81c4 (Metraipolis archive dc_ef07069c81c4, sha256 8d294334ee8e5bf0844b5a78de7bb5fc107e347aa331db557492660cee4f187f)
BibTeX
@misc{dcef07069c81c4,
  title = {Signatures of semantic search in the activations of large language models — arXiv:2609.35599},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35599},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ef07069c81c4; sha256 8d294334ee8e5bf0844b5a78de7bb5fc107e347aa331db557492660cee4f187f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35599. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

When recalling lists of concepts (e.g., animals) during the semantic fluency task (SFT), both humans and large language models (LLMs) organise their output into clusters of related items (e.g., sea animals) that are punctuated by strategic switches between clusters. In humans, this pattern can be explained by a semantic foraging process, whereby distinct neural and behavioural signatures accompany within-cluster production ("exploit") and between-cluster switching ("explore"). Whether LLMs likewise represent these two search regimes within their internal states is unknown. Here, we apply a range of mechanistic interpretability techniques to provide evidence for this. In Study 1, we use the Jacobian lens (J-lens), which maps intermediate-layer residual-stream representations to token-level activations, to show that concept-level activations predict switching. First, we find that switching coincides with low next-token activations. Moreover, the probability of switching rises as the set of strongest J-lens activations (the J-space) becomes depleted of items from the category currently being produced, analogous to explore-exploit decision-making during patch foraging. We then show that middle-layer J-lens activations of abstract category-related labels (e.g., "water") increase in anticipation of switching into that category. We confirm these representations to causally influence switching by deriving steering vectors that target category switching. In Study 2, we identify generic residual stream directions that are activated during and in anticipation of switching. By steering activations along these directions, we bias increased or decreased rates of switching. Our study extends the semantic foraging framework to artificial intelligences and provides evidence that LLMs maintain distinct representational signatures for exploration and exploitation as they verbalise conceptual information.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.35599 — 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 2026-10-07T20:29:01.905Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).