Zhihao Wang王志浩
Menu

News/September 1, 2026

NIKA public leaderboard and nika-bench 0.2.0

nika-bench 0.2.0 freezes Dev/Test splits for official evaluation. Leaderboard scoring uses rule-based RCA F1; submissions can pack trajectories and open a Hugging Face dataset PR.

The leaderboard site ranks agent performance under the frozen release. Earlier, nika-bench 0.1.0 and the submit pipeline landed in July 2026.