Computer Science & AI · arXiv:2610.22362
CoDA: Co-Evolution of Data Engines and Action Policies
Authors: Tianjiao Yu, Zhilin Ou, Yifan Shen, Xinzhuo Li, Yuanzhe Liu et al. (arXiv:2610.22362) · arXiv: 2610.22362
Core Methodological Innovation
Featured in Nujan's Computer Science & AI Literature Digest. Open side-by-side with interactive equation, methodology, and citation breakdowns.
Key Quantitative & Theoretical Takeaway: Open arXiv:2610.22362 in Nujan to inspect every mathematical symbol, experimental baseline, and citation directly alongside the original PDF.
Abstract
Generative simulation can automate robot environment construction, but physically valid environments are not necessarily useful for training. The challenge is to identify environments that address a policy's failures and supervision that helps it overcome them. We introduce CoDA, a framework in which a simulation data engine and an action policy co-evolve through a shared heterogeneous memory of environment construction, policy failures, and learning progress. The memory compiles validated diagnoses from a vision-language critic into generation programs that recreate difficulties and training programs that help the policy overcome them. We further propose Text2potential to convert these diagnoses into staged potentials for action-chunk value learning, providing dense supervision for policy updates while preserving the task's optimal policy under standard shaping assumptions. Measured learning progress guides further generation and program refinement. Across four manipulation task families, CoDA improves out-of-distribution success from 5.3% after supervised fine-tuning to 27.6%, exceeding the strongest closed-loop baseline by 9.5 percentage points with the same generation budget. On LIBERO-Pro, 15 rounds of generation and training raise success from 25% to 84% under position perturbations and from 1% to 76% under instruction perturbations.
Step-by-Step Equation & Methodology Breakdown
What is CoDA: Co-Evolution of Data Engines and Action Policies (arXiv:2610.22362) about?
Generative simulation can automate robot environment construction, but physically valid environments are not necessarily useful for training. The challenge is to identify environments that address a policy's failures and supervision that helps it overcome them. We introduce CoDA, a framework in which a simulation data engine and an action policy co-evolve through a shared heterogeneous memory of environment construction, policy failures, and learning progress. The memory compiles validated diagnoses from a vision-language critic into generation programs that recreate difficulties and training programs that help the policy overcome them. We further propose Text2potential to convert these diagnoses into staged potentials for action-chunk value learning, providing dense supervision for policy updates while preserving the task's optimal policy under standard shaping assumptions. Measured learning progress guides further generation and program refinement. Across four manipulation task families, CoDA improves out-of-distribution success from 5.3% after supervised fine-tuning to 27.6%, exceeding the strongest closed-loop baseline by 9.5 percentage points with the same generation budget. On LIBERO-Pro, 15 rounds of generation and training raise success from 25% to 84% under position perturbations and from 1% to 76% under instruction perturbations.
How can I read arXiv:2610.22362 with interactive explanations?
Visit https://nujan.app/abs/2610.22362 to open CoDA: Co-Evolution of Data Engines and Action Policies in Nujan's split-screen research workspace with instant equation and methodology breakdowns.