General Tabular Reasoning on MultiHiertt (dev)
71.17AccuracyMOCA-AGENT
Evaluation results
| Method | Links | |
|---|---|---|
| MOCA-AGENTMethod Category=Reinforcement learning, Backbone=Qwen3.6-27B2026.06 | 71.17 | |
| RL w/ CS (Formula)Method Category=Reinforcement learning2026.06 | 56.78 | |
| NAPGMethod Category=Fine-tuning approaches2026.06 | 44.19 |