Enhance Env step with adaptive penalties - #1

Open

chenmaike wants to merge 1 commit into

mainfrom

codex/update-dvp-penalty-to-adaptive-method

chenmaike commented Dec 18, 2025

Owner

Summary

add adaptive Lagrangian penalty with soft barrier for DVP constraint violations
tighten infeasible action handling using candidate mask-aware penalties
refresh environment resets to keep candidate masks and lambda state consistent

Testing

python -m compileall test.py


          Implement adaptive DVP penalty

d60708d

chenmaike added the codex label

— with

ChatGPT Codex Connector

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels