deep-learningarxiv

CARE: Grounded Recursive Self-Improvement for Production Customer-Service Agents

Recursive self-improvement (RSI) in production agents involves transforming accumulated experience into reliable model updates while adapting to evolving business policies. We present CARE…

Yanbei Jiang / Chao Lei / Rena Gao / Xiaochen Zhao / Yuchen Ge

用途
原典で確認
難易度の目安
Hard
実装コストの目安
High

概要

Recursive self-improvement (RSI) in production agents involves transforming accumulated experience into reliable model updates while adapting to evolving business policies. We present CARE (Customer-service Agent Retrospective and Evolve), a grounded RSI framework that operationalizes this process in customer service. Its goal is to replace a frontier proprietary incumbent with an in-house mid-sized model that continues learning after deployment. After a one-time cold start on governed production traces, the current model serves production traffic alongside the incumbent, and its own failures drive the next update. A 24/7 inspection process establishes failures through evidence-blinded, independently reviewed judgments. Attribution separates model errors from defects in Skills, tools, or the harness, routing component failures to repair. Model-attributable trajectories are truncated at the first erroneous decision and paired with verified repairs. On-policy self-distillation (OPSD) then updates the model using the same model as teacher, with the diagnosis and repair supplied as privileged information unavailable to the student. To support continued evolution as business requirements change, training mixes failures from different Skill/SOP versions, each judged against the policy in force when it occurred. Candidates run an executable simulation test compiled from the latest SOP and then frozen, followed by offline inspection of the resulting traces. Promotion requires a combined simulation-and-inspection score at least at parity with the deployed version, confirmation through an online comparison with the incumbent, and human approval.

参考リンク

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
CARE: Grounded Recursive Self-Improvement for Production Customer-Service Agents

# 概要
Recursive self-improvement (RSI) in production agents involves transforming accumulated experience into reliable model updates while adapting to evolving business policies. We present CARE…

# 原典
https://arxiv.org/abs/2611.02839

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。

関連する記事

github実装リンクあり

FastGen — NVIDIA FastGen: Fast Generation from Diffusion Models

この論文では、ディフュージョンモデルの高速化を目的としたNVIDIA FastGenについて説明しています。FastGenは、ディフュージョンモデルから高速に生成することが可能です。

ディフュージョンモデルの高速化