arXiv 2609.30199 · cs.AI · 2026-09-24
ExplorationBench: Measuring AI Systems' Exploration in Verifiable Alien Worlds
Ming Zhang, Zhenghao Xiang, Peizhong Gao, Yujiong Shen, Yuhui Wang, Zhonghan Yue, Shihan Dou, Zhangyue Yin 等 20 人
Tencent Hunyuan
在最新榜 #237 本站收录于 2026-10-21
一句话要点
腾讯混元 ExplorationBench 在可验证的虚构世界中测 AI 的科学探索能力，排除预训练知识的干扰。
Tencent Hunyuan's ExplorationBench measures scientific exploration in verifiable alien worlds, ruling out recall from pre-training.
阅读 arXiv 原文 PDF Hugging Face 讨论 代码仓库 项目页
主题与学科
同主题：AI for Science
- Scaling Parameter-Efficient Fine-Tuning for Brain Encoding ▲237
- EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery ▲112
- Predictive Credit: Measuring What Scientific Explanations Add to Experimental Forecasts ▲103
- PDE-JEPA: Predictive Representation Learning of Latent Dynamics Modeling for Parametric PDEs ▲65
- From Research Goals to Scientific Closure: Evaluating Agents Through Evidence-Supported Physics Research ▲53
数据说明
元数据来自 arXiv API（已核实）；点赞来自 Hugging Face 每日论文；引用来自 Semantic Scholar。最近刷新 2026-10-22。
要点由本站根据摘要自写，不转载摘要原文，请阅读原文。