用竞赛式评估让AI教师互评，再合作构建可验证的编程训练体系。
Compete Then Collaborate: Frontier AI Teachers Build a Verifiable Curriculum to Improve a Coding Student Beyond Imitation
- 四名顶级AI教师互相比拼，用执行测试公平打分
- 学生在可验证环境中强化学习，竞赛题正确率提升49%
- 适合想突破模仿学习瓶颈的模型训练者
大型语言模型日益作为教师为小型学生生成训练数据。现有方法合并多个教师输出但无法判断谁更优，常依赖自身偏好的LLM评判器。本文提出‘竞争后协作’框架：四名前沿教师（Claude、Codex-GPT、Grok、Gemini）在执行验证（单元测试与输入输出检查）下相互比拼，并通过公平控制排序；随后协同构建可验证的训练课程供学生（Qwen2.5-Coder）使用。研究发现：(1) 在执行验证下，所有教师自修正后标准问题解决率达99-100%，但难题区分度明显（Gemini 77% > Claude 69% = Codex 69% > Grok 50%）；然而学生表现不依赖教师排名。(2) 对已具备能力的学生进行模仿学习（SFT）反而降低性能（如在MBPP-test上从76.7%降至72.7%，竞赛题从5.9%降至2.9%）。(3) 同一协作课程作为强化学习带可验证奖励（RLVR）环境时，学生表现显著提升（竞赛题正确率从5.9%升至8.8%，相对增益49%），逆转了SFT的负面效果。核心价值在于共同构建可验证的学习环境，使学生通过实践而非模仿成长。项目提供可在NVIDIA GB10部署的可复现本地化流水线及GRPO框架补丁。
原文摘要 · Abstract (English)
Large language models increasingly serve as teachers generating training data for smaller students. Prior multi-teacher knowledge distillation methods merge outputs without determining which frontier model teaches best, often relying on an LLM judge biased toward its own outputs. We introduce a compete-then-collaborate framework where four frontier AI teachers (Claude, Codex-GPT, Grok, Gemini) are ranked head-to-head by an execution-based judge (unit tests and stdin-stdout checks) with fairness controls, and then collaborate to build a verifiable curriculum for a student (Qwen2.5-Coder). We report three findings. (1) Under execution verification, all teachers solve standard problems near-perfectly after self-correction (99-100%) due to a saturation effect, but harder competition problems separate them (Gemini 77% > Claude 69% = Codex 69% > Grok 50%); however, the robust student-side results do not depend on teacher ranking. (2) Imitation (SFT) on verified solutions does not improve, and can degrade, an already-competent student at 7B and 32B (e.g., from 76.7% to 72.7% on MBPP-test, and 5.9% to 2.9% on competition problems). (3) Using the same collaborative curriculum as a reinforcement learning with verifiable rewards (RLVR) environment improves the student (from 5.9% to 8.8% peak on competition problems, a +49% relative gain), reversing SFT's direction. The value of AI-teacher collaboration lies not in pooling answers to imitate, but in jointly constructing a verifiable environment where the student learns by doing. We release a reproducible on-prem pipeline (NVIDIA GB10) with framework patches for running GRPO on a bleeding-edge stack.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。