OGLS-SD improves on-policy self-distillation stability and math reasoning performance by constructing an outcome-discriminative steering direction from contrasts between successful and failed teacher logits.
A lign D istil: Token-Level Language Model Alignment as Adaptive Policy Distillation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
No citing papers match the current filters.