BabelJudge: Measuring LLM-as-a-Judge Reliability Across Languages and Agent Trajectories

23 June 2026 at 06:00 CEST