SciCode wrongly rejects correct solutions, a user says, citing an audit
A user quotes a domain-expert audit of all 65 SciCode test problems that reports 263 defects.
TLDR
A user challenges SciCode’s reliability, quoting an audit that reports 263 defects in a review of all 65 test problems. According to the quoted passage, 192 defects, spread across 91% of the main problems, cause correct, instruction-following solutions to be wrongly rejected.
Combined views
10.4K
1 Source, first seen 27d ago
likes