Spaces:
Running
Running
Commit History
-
Add semantic 'Why' metric, redesigned hard task, multi-model benchmarker, and findings paper
7f5e6a9 -
Update tests and reward print formatting
Align tests with updated scoring and reward behavior and adjust console output formatting. Tests in test_advanced_cases.py, test_comprehensive.py, test_environment.py, test_graders.py, and test_rewards.py were updated to expect the new minimal score floor (0.001), adjusted reward values (e.g. false-positive/duplicate handling returning 0.01), and revised final/efficiency score expectations (e.g. 0.999 / 1.099). Also change inference._print_end to format individual rewards with three decimal places for consistent output.7189ec7
-
Final change
4a93440 -
chagne
4107f0a
-
removal
57a9d33 -
v1.1.1 final
748cbbc -
v1.0.0
f9fa317