pino-source-code / data /benchmarks

Commit History

chore: READY gate state, material_profiles v11_5 refresh, source-population AI review, repair-pairs benchmark
fe0bb36

Matthew Ford commited on

feat: expand substitution-triplet benchmark 8 -> 20
e6d699c

Matthew Ford commited on

feat: prospective formula benchmark (40 in-silico formulas, blinded labels)
3fb95e8

Matthew Ford commited on

feat: complete representation ablation + discordant substitution triplets
45a164d

Matthew Ford commited on

feat: freeze representation validation gate
2ff3288

Matthew Ford commited on