Metraipolis › archive › dc_5227508dcfbd
ACCT: Asymmetric Cross-LLM Context Transfer — arXiv:2610.42116
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.42116
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T07:32:24.672Z
- SHA-256 of the archived text
46966e7c031a6053fb67e87e63b8553a540c94b8efb3a5a827b0bef80f04b1ca- Stable URL of this page
https://metraipolis.com/documento/dc_5227508dcfbd
How to cite
ACCT: Asymmetric Cross-LLM Context Transfer — arXiv:2610.42116 — arXiv (en), https://arxiv.org/abs/2610.42116 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5227508dcfbd (Metraipolis archive dc_5227508dcfbd, sha256 46966e7c031a6053fb67e87e63b8553a540c94b8efb3a5a827b0bef80f04b1ca)
BibTeX
@misc{dc5227508dcfbd,
title = {ACCT: Asymmetric Cross-LLM Context Transfer — arXiv:2610.42116},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.42116},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5227508dcfbd; sha256 46966e7c031a6053fb67e87e63b8553a540c94b8efb3a5a827b0bef80f04b1ca},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.42116. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
In real-world deployments, large language model (LLM) systems are increasingly evolving from single-model invocation toward heterogeneous deployments in which different models are dynamically invoked for inference. When the same context is reused across models, however, each newly selected target model typically has to process it again, repeatedly incurring substantial computational costs. Therefore, we propose ACCT (Asymmetric Cross-LLM Context Transfer), which targets KV cache transfer between models of different scales within the same model family. ACCT transforms the KV cache produced by a source model into sparse KV states that can be directly consumed by a target model, thereby enabling cross-model reuse of context computation and decoupling context computation from downstream inference. ACCT consists of three lightweight components: the Cartographer identifies query-relevant context tokens for each target layer and KV head; the Allocator determines how many context tokens each target layer and KV head retains under a shared total KV budget; and the Compiler transforms selected source KV into sparse target KV. The target model can therefore perform inference using the transferred KV without reprocessing the original context. We evaluate ACCT in two complementary settings comprising five general-capability benchmarks and six highly context-dependent tasks. Experiments show that ACCT matches or exceeds the target model's original inference performance even when reducing the KV cache size by 75%. At a context length of approximately 19K tokens, ACCT's KV conversion achieves a 6.5-21.2 times speedup over target context prefill. These results demonstrate the feasibility of transferring context computation across models under a limited KV budget. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.42116 — 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 2026-11-03T07:29:33.665Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).