Metraipolis › archive › dc_67daba5abfa3

Beyond the Opening: Tracing the First Token's Persistent Influence on Reasoning Length — arXiv:2610.34350

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.34350
License (as archived): CC0-1.0
Archived: 2026-10-27T08:11:34.306Z
SHA-256 of the archived text: 02e43e28be1dd02bfd6f08b7659b64056659c18c3714bf7729fd34496b463ce6
Stable URL of this page: https://metraipolis.com/documento/dc_67daba5abfa3

How to cite

Beyond the Opening: Tracing the First Token's Persistent Influence on Reasoning Length — arXiv:2610.34350 — arXiv (en), https://arxiv.org/abs/2610.34350 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_67daba5abfa3 (Metraipolis archive dc_67daba5abfa3, sha256 02e43e28be1dd02bfd6f08b7659b64056659c18c3714bf7729fd34496b463ce6)

BibTeX

@misc{dc67daba5abfa3,
  title = {Beyond the Opening: Tracing the First Token's Persistent Influence on Reasoning Length — arXiv:2610.34350},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.34350},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_67daba5abfa3; sha256 02e43e28be1dd02bfd6f08b7659b64056659c18c3714bf7729fd34496b463ce6},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34350. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Understanding how individual token choices affect reasoning length is important for characterizing reasoning models. Building on prior findings that changing the first reasoning token (hereafter, the first token) can substantially alter reasoning length, we focus on this initial choice to investigate how its influence operates during subsequent generation. One possibility is that the first token changes the early text that follows, which in turn affects reasoning length. However, replacing only the first token of an existing reasoning prefix and resuming generation substantially changes the remaining reasoning length, even though all other prefix tokens are unchanged. To trace this persistent influence, we compare next-token distributions while varying only the first token and find that sensitivity is concentrated at a small set of generation positions. Blocking attention from these positions to the first token reduces the length gap between first-token conditions. Using next-token distributions from a different first-token condition at these positions also substantially changes subsequent reasoning length. Together, these findings suggest that attention to the first token at a few sensitive positions helps sustain its length effect, illustrating how a local token choice persistently influences a sequence-level outcome.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.34350 — 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 2026-10-27T07:56:00.228Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).