Metraipolis › archive › dc_95e5d2fec216
The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents — arXiv:2609.20050
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.20050
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T03:24:01.153Z
- SHA-256 of the archived text
581c908108749c71fe2f6a8cf588db698222417fa8449d0e8b5f30e03a531ad2- Stable URL of this page
https://metraipolis.com/documento/dc_95e5d2fec216
How to cite
The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents — arXiv:2609.20050 — arXiv (en), https://arxiv.org/abs/2609.20050 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_95e5d2fec216 (Metraipolis archive dc_95e5d2fec216, sha256 581c908108749c71fe2f6a8cf588db698222417fa8449d0e8b5f30e03a531ad2)
BibTeX
@misc{dc95e5d2fec216,
title = {The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents — arXiv:2609.20050},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.20050},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_95e5d2fec216; sha256 581c908108749c71fe2f6a8cf588db698222417fa8449d0e8b5f30e03a531ad2},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.20050. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Retrieval assembles repository context by ranking passages for relevance to the current query. A coding agent halfway through an issue has already read much of what such a ranker returns. Relevance is scored per passage, but sufficiency belongs to the set: independently scored passages can fill the budget with support for one requirement while another goes unmet. We formulate state-conditioned minimal sufficient evidence recovery: given a captured agent state, recover a compact evidence combination supplying what its next decision still lacks. SERBench measures this on 500 held-out states from 45 repositories, recording what the agent has seen, crediting only sets that satisfy every annotated evidence requirement of the current decision, and separating set recovery from candidate discovery. MSS-Complement treats acquisition as set construction, not ranking. Three semantic calls propose a jointly sufficient set, search for what it lacks, and return 4-8 intact source units within 6,144 tokens. One configuration, fixed on calibration data, recovers a complete set for 73.0% of those states at five items and 80.6% at eight, against 61.4% and 72.4% for Qwen3 embedding with reranking. A matched control ranking by similarity alone recovers fewer complete sets, placing the margin over it in the set-level policy, not the computation. The lead persists from frozen repository source with no gold-derived pool. On AMA-Bench it answers from a 76.2% smaller answer prompt, with accuracy 2.08 points above that benchmark's own memory agent. Removing one required group from a complete set costs repair-localization precision under two executors. Retrieval for agents is better posed as recovering what a decision lacks than re-ranking what an issue resembles. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.AI). Categorie arXiv: Information Retrieval (cs.IR) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.20050 — 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 2026-10-10T03:04:13.892Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).