last commit

Add detailed model performance reasoning across all benchmark documentation

Update docs with latest HF Native and OpenRouter benchmark scores

Compliance fix: Move inference.py to repo root strictly enforcing OpenEnv hackathon submission rules

Final cleanup: Remove redundant testing scripts, un-track logs, sanitize comments

Fix NoneType subscript bug in json parser pipeline

Add extreme final submission tests (48 tests: math, load, cross-file, adversarial, compliance)

v2.0.1

fix: clamp inference score strictly to 0.999 to avoid float format rounding to 1.000

changes

experimental