Science Multiple-Choice Question Answering on ARC Challenge
91AccuracyBerr. Text
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Berr. TextCollaboration space=Text-space, Description=Best text-space collaboration result2026.02 | 91 | — | |
| Berr. BLDescription=LLaDA-only baseline by Berrayana et al.2026.02 | 87.5 | — | |
| Berr. LatentCollaboration space=Latent-space, Description=Best latent-space DDLM→ARM result, Projector training=trained Linear-GELU-Linear projector on 35K domain-specific samples2026.02 | 81 | — | |
| LLaDA + Sonnet PlanPlan conditioning=Sonnet Plan2026.02 | 75.2 | 3.9 | |
| LLaDADescription=LLaDA-only baseline2026.02 | 71.3 | — |