Metraipolis › archive › dc_38280b951b0a

On the Opportunities and Challenges of Prefill-to-Autoregressive Cross-Model Transfer — arXiv:2611.06976

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.06976
License (as archived)
CC0-1.0
Archived
2026-11-06T14:43:00.002Z
SHA-256 of the archived text
d34b76f2f98cdc20cae91cff7881d3d559d1127e110ceacd3c42c725c2e5e4c0
Stable URL of this page
https://metraipolis.com/documento/dc_38280b951b0a

How to cite

On the Opportunities and Challenges of Prefill-to-Autoregressive Cross-Model Transfer — arXiv:2611.06976 — arXiv (en), https://arxiv.org/abs/2611.06976 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_38280b951b0a (Metraipolis archive dc_38280b951b0a, sha256 d34b76f2f98cdc20cae91cff7881d3d559d1127e110ceacd3c42c725c2e5e4c0)
BibTeX
@misc{dc38280b951b0a,
  title = {On the Opportunities and Challenges of Prefill-to-Autoregressive Cross-Model Transfer — arXiv:2611.06976},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.06976},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_38280b951b0a; sha256 d34b76f2f98cdc20cae91cff7881d3d559d1127e110ceacd3c42c725c2e5e4c0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.06976. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Prefill-to-Autoregressive Cross-model Transfer (PACT), a form of latent communication between language models, combines a larger model's prompt representations with a smaller model's efficient generation, offering a promising way to balance capability and inference cost. However, existing systems provide limited support for independently evaluating large-small model collaboration, and improved task performance alone does not establish whether the receiver benefits from the sender's information. We introduce PLatour, an open-source PACT framework that connects frozen models through lightweight, zero-initialized linear bridges trained directly with task rewards and supports hybrid-attention architectures. PLatour significantly outperforms open-source baselines on generation tasks and surpasses Mostik's reported results on shared mathematics and coding benchmarks using a substantially smaller sender. However, controlled experiments show that the gains on some tasks can be reproduced by uninformative sender signals, self-communication, or receiver-only training. We find that trainable bridges can act as parameter-efficient adaptation mechanisms without exploiting useful sender information. We therefore propose PACT-Check to quantify how much of a trained PACT system's gain genuinely comes from communication. Finally, we discuss the opportunities and challenges of PACT systems. Code is available at this https URL .

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/2611.06976 — 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 2026-11-06T14:18:57.580Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).