TuringSpace用交互式验证码测试空间智能
AI 总结
Xuwei Ding等12名作者推出TuringSpace,这是一个包含125项交互式验证码任务、覆盖五类空间能力的合成网页环境。论文称,人类的任务成功率为94.1%,而GPT-6 Astra的失败率为25.9%,Kimi K3的失败率为68.0%。在该环境中微调Qwen3.5-9B后,其成功率从1.6%升至33.1%,并在七个分布外基准上均有提升。
深度见解
该预印本提供了一套可交互的评测与训练环境,用于考察多模态模型能否感知、行动并修正空间决策,且报告的提升可迁移到真实场景。
Computer Science > Machine Learning
arXiv:2611.25387 (cs)
[Submitted on 19 Nov 2026]
Title:TuringSpace: Grounding Spatial Intelligence in Action with Interactive CAPTCHAs
Authors:Xuwei Ding, Skylar Zhai, Linxin Song, Jiarui Feng, Dongqi Fu, Huajun Xi, Junnan Li, Jingcheng Liang, Jieyu Zhao, Yinglong Xia, Hong Li, Jian Kang
View a PDF of the paper titled TuringSpace: Grounding Spatial Intelligence in Action with Interactive CAPTCHAs, by Xuwei Ding and 11 other authors