Static Evaluation of Model Switching in LLM Agents Scores the Wrong Worldarxiv.org4 points·danbitengo··0 commentsOpen articleSaveView on HN