ResearchBenchmarksTable Reasoning on MHFollow47.7Pass@1TAC2.4614.20525.9537.695Jun 23, 2026Evaluation ResultsMethodMethodLinksPass@1TACBackbone=Qwen3-4B-Base...Backbone=Qwen3-4B-Base, Seeds=12026.0647.7SECBackbone=Qwen3-4B-Base...Backbone=Qwen3-4B-Base, Seeds=12026.0645.4RandomBackbone=Qwen3-4B-Base...Backbone=Qwen3-4B-Base, Seeds=12026.0644.9TACBackbone=Qwen3-0.6B-Ba...Backbone=Qwen3-0.6B-Base, Seeds=12026.0612.3SECBackbone=Qwen3-0.6B-Ba...Backbone=Qwen3-0.6B-Base, Seeds=12026.0612RandomBackbone=Qwen3-0.6B-Ba...Backbone=Qwen3-0.6B-Base, Seeds=12026.064.2