We have 2 of 7 papers
We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.
Not the right person? Other researchers publish under this name.
The same recipe trained on a multi-domain corpus improves the frozen backbone on twelve evaluation benchmarks spanning mathematics, science, and commonsense, without task-specific adaptation.
RG-OPD bridges sparse verifier rewards and dense teacher logits, preserving token-level supervision while filtering misleading teacher signals, and produces stronger distilled students, outperforming both vanilla reverse-KL distillation and the recent TSD-KD baseline.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
