Fix reward always in open interval (0,1) for Phase 2 validation e893f88 Shubham-Rasal commited on Apr 8
fix(Phase 2): restore HF Space YAML header + reward Pydantic default to 1e-4 648a438 Shubham-Rasal commited on Apr 8
fix(Phase 2): map all observation rewards to open interval (0,1) b88fbfe Shubham-Rasal commited on Apr 8
fix: map grader scores to open interval (0,1) for Phase 2 validation 5d242e2 Shubham-Rasal commited on Apr 7