arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.30116cs.CLcs.AI
收藏

环境对决:大语言模型协同演化的微观世界

Environment Duels: A Microcosm of LLM Co-Evolution

University of Pennsylvania(宾夕法尼亚大学)

Shibo Jin, Zhiqiu Xu, Yiling Zou, Shangyu Gong, Mayur Naik

AI总结:

本研究提出动态智能体基准Environment Duels,评估9个前沿LLM的求解与创作能力,发现二者为不同技能,且模型交互模式可反映LLM协同演化规律。

摘要:

通过从合成训练管道到开放网络共享数据等一系列现实世界交互,前沿大语言模型(LLM)会吸收其前身与竞争对手所创造的内容。每个模型在这个生态系统中作为创造者和求解者的表现如何?我们提出了Environment Duels,这是一个动态智能体基准,用于评估模型在交互式环境中的两种能力。每个模型编写环境,每个环境都带有一个特权提示,目标是让有能力的求解者在没有该提示时无法完成任务,而有该提示时可以完成。在9个前沿LLM中,求解和创作是不同的技能:最强的求解者并非最强的创作者,而较小的模型能从其他模型的交互方式中学到更多,写出最难的环境和最有用的提示。除了分数外,成对交互形成了模型相互影响的微观世界:头对头结果是可传递的,因此能产生稳定排名;独立创作者会趋同于相同的设计;在使用Qwen3.8-27B的强化学习研究中,学习者在作者环境上的提示提升,与在这些环境上的训练对其在保留的推理和智能体基准上的提升程度相关,无论这些环境是其自身的还是其他模型的。这些模式都与模型在对决之外相互塑造的方式相对应。通过在受控条件下上演这些交互,Environment Duels为LLM如何相互教导、挑战和学习提供了一个窗口。代码可在该https URL获取。

英文摘要:

Through a tangle of real-world interactions, from synthetic training pipelines to data shared on the open web, frontier LLMs absorb what their predecessors and competitors create. How well does each model play its role in this ecosystem, as a creator and as a solver? We introduce Environment Duels, a dynamic agentic benchmark that evaluates both capabilities on interactive environments. Each model writes environments, each with a privileged hint, aiming for tasks that capable solvers fail without the hint and pass with it. Across nine frontier LLMs, solving and authoring prove to be distinct skills: the strongest solvers are not the strongest authors, while smaller models learn more from how others played and write the hardest environments and the most useful hints. Beyond the scores, the pairwise interactions form a microcosm of how models affect one another. Head-to-head outcomes are transitive and so yield a stable ranking; independent authors converge on the same designs; and in a reinforcement-learning study with Qwen3.8-27B, the learner's hint lift on an author's environments tracks how much training on those environments improves it on held-out reasoning and agentic benchmarks, whether the environments are its own or another model's. Each of these patterns has a counterpart in how models shape one another outside the duel. By staging these interactions under controlled conditions, Environment Duels offers a window onto how LLMs teach, challenge and learn from one another. Code is available at https://github.com/LLM-Duels/EnvDuels.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑