Metraipolis › archive › dc_de83a75139fe

HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning — arXiv:2610.03039

Source
arXiv
Original
https://arxiv.org/abs/2610.03039
License (as archived)
CC0-1.0
Archived
2026-10-05T22:51:40.782Z
SHA-256 of the archived text
08f260e51418cf58b58bbe7131eea51290902518ae4396c9b694d86e93310076
Stable URL of this page
https://metraipolis.com/documento/dc_de83a75139fe

How to cite

HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning — arXiv:2610.03039 — arXiv, https://arxiv.org/abs/2610.03039 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_de83a75139fe (Metraipolis archive dc_de83a75139fe, sha256 08f260e51418cf58b58bbe7131eea51290902518ae4396c9b694d86e93310076)
BibTeX
@misc{dcde83a75139fe,
  title = {HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning — arXiv:2610.03039},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.03039},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_de83a75139fe; sha256 08f260e51418cf58b58bbe7131eea51290902518ae4396c9b694d86e93310076},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.03039. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-form thinking traces can substantially improve the multi-step reasoning performance of large language models (LLMs), but they introduce high inference-time overhead, with latency dominated by sequential decoding. We propose HyperThink, a text-to-parameter approach that amortizes this reasoning computation into a single query-conditioned parameter update: a lightweight hypernetwork reads the question and predicts updates to a small subset of the base LLM's parameters, while a vector-quantized decoder constrains them to a finite set of reusable patterns to improve robustness and transfer. Trained end-to-end on outputs from the base model itself, HyperThink eliminates long thinking traces at test time: after one hypernetwork forward pass, the adapted model generates a concise step-by-step solution and final answer without an intermediate trace, using far fewer tokens while retaining strong reasoning performance. Empirically, HyperThink improves the low-latency region of the accuracy-latency trade-off on mathematical and general reasoning tasks, with its strongest gains in the near-non-thinking regime.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.03039 — 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 2026-10-05T22:46:10.566Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).