WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild

2025-01-22 · ICLR 2025 Spotlight · anchor

Findings