亲爱的研友该休息了!由于当前在线用户较少,发布求助请尽量完整地填写文献信息,科研通机器人24小时在线,伴您度过漫漫科研夜!身体可是革命的本钱,早点休息,好梦!

CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation

计算机科学 机器人 可执行文件 具身认知 稳健性(进化) 人工智能 标杆管理 感知 编码(社会科学) 人机交互 强化学习 内含代理 机器人学 概率逻辑 渲染(计算机图形) 仿人机器人 编码器 机器学习 可验证秘密共享 自主代理人 控制(管理) 可靠性(半导体) 视觉伺服 主动感知 冗余(工程) 先验概率 认知机器人学 演示式编程 瓶颈
作者
Max Fu,Justin Yu,Karim El-Refai,Ethan Kou,Haoru Xue,Huang Huang,Wenli Xiao,Guanzhi Wang,Fei-Fei Li,Guanya Shi,Jiajun Wu,Shankar Sastry,Yuke Zhu,Ken Goldberg,Linxi Fan
出处
期刊:Cornell University - arXiv [Cornell University]
摘要

"Code-as-Policy" considers how executable code can complement data-intensive Vision-Language-Action (VLA) methods, yet their effectiveness as autonomous controllers for embodied manipulation remains underexplored. We present CaP-X, an open-access framework for systematically studying Code-as-Policy agents in robot manipulation. At its core is CaP-Gym, an interactive environment in which agents control robots by synthesizing and executing programs that compose perception and control primitives. Building on this foundation, CaP-Bench evaluates frontier language and vision-language models across varying levels of abstraction, interaction, and perceptual grounding. Across 12 models, CaP-Bench reveals a consistent trend: performance improves with human-crafted abstractions but degrades as these priors are removed, exposing a dependence on designer scaffolding. At the same time, we observe that this gap can be mitigated through scaling agentic test-time computation--through multi-turn interaction, structured execution feedback, visual differencing, automatic skill synthesis, and ensembled reasoning--substantially improves robustness even when agents operate over low-level primitives. These findings allow us to derive CaP-Agent0, a training-free framework that recovers human-level reliability on several manipulation tasks in simulation and on real embodiments. We further introduce CaP-RL, showing reinforcement learning with verifiable rewards improves success rates and transfers from sim2real with minimal gap. Together, CaP-X provides a principled, open-access platform for advancing embodied coding agents.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
罗Eason应助爱听歌的悒采纳,获得30
刚刚
完美世界应助爱听歌的悒采纳,获得30
刚刚
星辰大海应助爱听歌的悒采纳,获得20
刚刚
酷波er应助爱听歌的悒采纳,获得10
刚刚
在水一方应助爱听歌的悒采纳,获得10
1秒前
JamesPei应助爱听歌的悒采纳,获得10
1秒前
欣欣发布了新的文献求助10
1秒前
molihuakai应助爱听歌的悒采纳,获得10
7秒前
Owen应助爱听歌的悒采纳,获得10
7秒前
大模型应助爱听歌的悒采纳,获得10
8秒前
Lucas应助爱听歌的悒采纳,获得10
8秒前
研友_VZG7GZ应助爱听歌的悒采纳,获得10
8秒前
8秒前
温柔的静丹完成签到,获得积分10
8秒前
研友_VZG7GZ应助爱听歌的悒采纳,获得10
9秒前
斯文败类应助爱听歌的悒采纳,获得10
9秒前
Jasper应助爱听歌的悒采纳,获得10
9秒前
小蘑菇应助爱听歌的悒采纳,获得10
9秒前
野性的幼萱完成签到,获得积分10
9秒前
沈轶烽完成签到,获得积分10
10秒前
沈轶烽发布了新的文献求助30
12秒前
Lucas应助爱听歌的悒采纳,获得20
16秒前
初景应助爱听歌的悒采纳,获得20
16秒前
传奇3应助爱听歌的悒采纳,获得30
17秒前
yao完成签到 ,获得积分10
17秒前
ding应助爱听歌的悒采纳,获得10
17秒前
科目三应助爱听歌的悒采纳,获得10
17秒前
脑洞疼应助爱听歌的悒采纳,获得10
17秒前
慕青应助爱听歌的悒采纳,获得10
18秒前
18秒前
CipherSage应助爱听歌的悒采纳,获得10
18秒前
所所应助爱听歌的悒采纳,获得10
18秒前
Julie完成签到,获得积分10
19秒前
李健的小迷弟应助沈轶烽采纳,获得10
23秒前
dongzh完成签到 ,获得积分20
24秒前
yao关注了科研通微信公众号
24秒前
24秒前
七慕凉发布了新的文献求助10
30秒前
科研通AI6.4应助笨笨静白采纳,获得10
33秒前
37秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Essentials of Carbohydrate Chemistry and Biochemistry, 4th Edition 800
Navigating Normative Orders. Interdisciplinary Perspectives 800
1 Peter and Christ's Descent to the Dead in Its Early Christian Reception 700
Organizational Behavior 510
Management and the Arts 510
Matrix Methods in Data Mining and Pattern Recognition Second Edition 510
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 工程类 有机化学 化学工程 生物化学 计算机科学 内科学 物理 复合材料 催化作用 细胞生物学 无机化学 光电子学 物理化学 电极 基因
热门帖子
关注 科研通微信公众号,转发送积分 7749790
求助须知:如何正确求助?哪些是违规求助? 9297533
关于积分的说明 20240651
捐赠科研通 7331159
什么是DOI,文献DOI怎么找? 3309381
关于科研通互助平台的介绍 2460916
邀请新用户注册赠送积分活动 2321673