List your company

Multi-Hop Question Answering on HotpotQA (Score)

38.9HotpotQA Scoreπ-Play
10.09217.57125.0532.529Apr 15, 2026

Evaluation results

MethodLinks
π-PlayBase Model=Qwen3-8B2026.0438.9
π-PlayBase Model=Qwen3-4B-Instruct-25072026.0438.5
Dr.ZeroBase Model=Qwen3-4B-Instruct-25072026.0438.4
Dr.ZeroBase Model=Qwen3-8B2026.0436.6
Search-R1Base Model=Qwen3-4B-Instruct-25072026.0434.9
SQLM*Base Model=Qwen3-4B-Instruct-25072026.0434.1
SQLM*Base Model=Qwen3-8B2026.0433.1
SQLM*Base Model=Qwen3-4B2026.0433
π-PlayBase Model=Qwen3-4B2026.0432.3
Search-R1Base Model=Qwen3-8B2026.0432
Dr.ZeroBase Model=Qwen3-4B2026.0431
Search-R1Base Model=Qwen3-4B2026.0430.7
ReActBase Model=Qwen3-4B-Instruct-25072026.0422.8
ReActBase Model=Qwen3-8B2026.0422.3
ReActBase Model=Qwen3-4B2026.0411.2