4 papers indexed here
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.
Expected Reasoning-Step Return is introduced, which treats semantic reasoning steps as macro-actions and uses Monte Carlo student-policy rollouts to estimate the expected final task reward of student-generated and teacher-proposed actions in a common return space for step-level comparison.
Calibrated On-Policy Distillation is introduced, which estimates the teacher's self-deviation region through positive and negative privileged interventions and calibrates the original teacher--student discrepancy by retaining only the component that lies beyond this region.
These findings show that privileged shifts fail to provide reliable answer-aligned directions, while their magnitudes primarily reflect counterfactual sensitivity rather than token-level learning value, and propose Counterfactual Sensitivity Credit Reallocation (CSCR), a simple extension of GRPO that reduces credit for...
This work presents MG2Act, a structure-independent framework that translates two-step logic into sequential cross-attention, using CRBN-mediated degradation as the most data-rich representative system.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
