CLAP: Closed-Loop Training, Evaluation, and Release Control for Domain Agent Post-training
Abstract
Domain agents often face noisy business data, uncertain post-training gains, offline/application mismatch, and adapter-release risk. This paper presents CLAP (Closed-Loop Agent Post-training), a closed-loop method that converts business data into structured SFT samples, decision-preference samples, holdout sets, risk diagnostics, and release-gate records. CLAP combines data validation, target/evidence normalization, reward/KL diagnosis, offline gates, and application-chain replay to decide whether an adapter is suitable for the target application chain. On five anonymized manufacturing-scenario batches, QLoRA-style LoRA-SFT yields modest average gains: overall score increases by 0.0098, pass rate by 0.0240, and evidence accuracy by 0.0280, while hallucination and wrong facts decrease. Yet only 3 of 5 batches improve, some batches regress, and GRPO exposes high KL risks. Application-chain replay further shows that RAG is necessary for factual extraction; under the same 3B backbone and 100 replay cases, an application-RAG-oriented LoRA-SFT adapter improves value, core fields, and answer-evidence doc/page matching over base+RAG, but increases latency. These results support managing domain-agent post-training through an integrated data-training-evaluation-release loop rather than relying on training completion or a single offline score.
Cite
@article{arxiv.2607.01846,
title = {CLAP: Closed-Loop Training, Evaluation, and Release Control for Domain Agent Post-training},
author = {Fangfei Li and Chenyang Zhao and Long Wang and Feng Tian and Zhiyue Zheng and Lv Guo},
journal= {arXiv preprint arXiv:2607.01846},
year = {2026}
}
Comments
6 pages, 1 figure. Accepted to CRAE 2026; to appear in SPIE Proceedings. Best Poster Award