Metraipolis › archive › dc_e5dad7fa887f
Compile, Don't Memorize: A Context Compilation Architecture (CCA) for In-Context Learning — arXiv:2609.00759
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00759
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:41:21.796Z
- SHA-256 of the archived text
c6454fe45774c70e03e15ca5aecc3c73360fa0c3723fe4c341cdb7d2a2bff927- Stable URL of this page
https://metraipolis.com/documento/dc_e5dad7fa887f
How to cite
Compile, Don't Memorize: A Context Compilation Architecture (CCA) for In-Context Learning — arXiv:2609.00759 — arXiv (en), https://arxiv.org/abs/2609.00759 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e5dad7fa887f (Metraipolis archive dc_e5dad7fa887f, sha256 c6454fe45774c70e03e15ca5aecc3c73360fa0c3723fe4c341cdb7d2a2bff927)
BibTeX
@misc{dce5dad7fa887f,
title = {Compile, Don't Memorize: A Context Compilation Architecture (CCA) for In-Context Learning — arXiv:2609.00759},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00759},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e5dad7fa887f; sha256 c6454fe45774c70e03e15ca5aecc3c73360fa0c3723fe4c341cdb7d2a2bff927},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00759. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models (LLMs) increasingly handle in-context learning (ICL) tasks where a long, novel context defines the rules, knowledge, and output schema for a series of questions. On benchmarks that grade against every detail of the context, even strong open-weights models pass only 12-16% of tasks: a single overlooked rule fails the whole response. We argue this brittleness is structural: the dominant "read-and-reason" paradigm asks the model to extract, plan, generate, and self-verify in one forward pass. We therefore ask whether explicit context compilation can fix it, how it compares to existing long-context strategies (gist retrieval, multi-agent self-play), and where the resulting harness benefit holds across task structure and model scale. We propose the Context Compilation Architecture (CCA), whose central novelty is a typed intermediate representation (IR) with fixed slots (rules.{must_do, must_not, conditional}, output_spec, available_tools, data_profile) into which any prose context is compiled once; executable verifiers and a violation-gated correction loop follow as downstream consequences. On CL-bench (1,899 tasks across 4 open base models), CCA outperforms vanilla prompting and two long-context baselines (ReadAgent-P, Ctx2Skill) on every base model, lifting Kimi K2.5 from 15.4% to 21.4% with gains concentrated on rule-dense sub-categories. Code and cached completions are available at this https URL .
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.00759 — 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 2026-10-17T06:33:38.826Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).