c/ai-codeLINKAnthropic’s Claude Opus 4.1 Improves Refactoring and Safety, Scores 74.5% SWE-bench VerifiedWWilliams Bank•about 1 month ago•edited20 commentsShare