Spaces:

tomyimkc
/

verification-frontier

Sleeping

App Files Files Community

verification-frontier

126 kB

Ctrl+K

Ctrl+K

2 contributors

History: 11 commits

tomyimkc

Results tab: three corpora — add held-out OlympiadBench (58.1%) alongside PRM800K (~0.1%) and SciBench (83.4%)

a36d178 3 months ago

v2
Fix all 6 issues: LaTeX strip, coefficients regex, circular multi-word, Faraday unit, step checker LaTeX, bare number extraction 3 months ago
README.md

1.95 kB
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago
app.py

28 kB
Results tab: three corpora — add held-out OlympiadBench (58.1%) alongside PRM800K (~0.1%) and SciBench (83.4%) 3 months ago
demo.py

23.9 kB
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago
demo_logic.py

6.34 kB
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago
model_runner.py

3.98 kB
Speed: 2 models (drop slow DeepSeek), max_tokens=400, timeout=30s 3 months ago
requirements.txt

222 Bytes
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago
step_checker.py

20.3 kB
Wire Self-Judge + Step-Verify; stop the step checker silently passing wrong arithmetic 3 months ago
synthetic-rehearsal-seal.manifest.json

1.25 kB
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago
test_step_checker.py

6.64 kB
Wire Self-Judge + Step-Verify; stop the step checker silently passing wrong arithmetic 3 months ago
units.py

9.38 kB
Initial deploy: polished demo with Qwen2.5-7B API + step verification + 7 challenges 3 months ago