Our survey on LLM agent trajectory analysis has been accepted for publication in IEEE Transactions on Software Engineering (TSE), providing a structured overview of how failures in increasingly autonomous AI agents can be attributed, analyzed, and used for system improvement.
[weiterlesen]