MEGA Hub

Fewer Tokens, Better Action: GPT-6 Astra Robot Agents with 14% Higher Success Rate but 65% Fewer Tokens

Authors

Do you know Ruiyang Si?You can claim authorship or link another user.Do you know Jianxin Bi?You can claim authorship or link another user.Do you know Shunyu Yang?You can claim authorship or link another user.Do you know Rui Ni?You can claim authorship or link another user.Do you know Wenbo Huang?You can claim authorship or link another user.Do you know Qiang Wang?You can claim authorship or link another user.Do you know Shulong Jiang?You can claim authorship or link another user.Do you know Duomin Wang?You can claim authorship or link another user.Do you know Xiuyu Li?You can claim authorship or link another user.Do you know Haiwen Feng?You can claim authorship or link another user.Do you know Zhen Dong?You can claim authorship or link another user.Do you know Daquan Zhou?You can claim authorship or link another user.

Abstract

Vision language model (VLM) agents can control robots through visual feedback and action primitives, but repeated model invocations and redundant observations incur substantial token overhead. We introduce PyRUA-Lean, an interactive code-execution framework that couples feedback-driven primitive composition with selective observation: the agent composes classical robot primitives and learned vision-language-action (VLA) policies into Python cells that perform conditional checks and local retries, returning only explicitly requested images and state feedback for replanning. Across 700 simulated task instances from LIBERO-PRO, RoboTwin 2.0, and RoboCasa365, we compare PyRUA-Lean with a tool-calling baseline using the same GPT-6 Astra planner and underlying robot primitives. Under equal LLM-call budgets, PyRUA-Lean increases overall success from 63.1% to 71.7%. On instances solved by both agents, it uses 49% fewer LLM calls and 65% fewer input tokens.

Community

00