pith. sign in

International Conference on Learning Representations , year =

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.LG 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

RASFT: Rollout-Adaptive Supervised Fine-Tuning for Reasoning

cs.LG · 2026-06-05 · unverdicted · novelty 6.0

RASFT is an adaptive SFT method that strengthens or relaxes expert imitation per problem based on on-policy rollout solvability and adds clipped reference-policy ratio to limit drift, reporting better results than standard SFT and RL on math and code benchmarks.

citing papers explorer

Showing 1 of 1 citing paper.

  • RASFT: Rollout-Adaptive Supervised Fine-Tuning for Reasoning cs.LG · 2026-06-05 · unverdicted · none · ref 2

    RASFT is an adaptive SFT method that strengthens or relaxes expert imitation per problem based on on-policy rollout solvability and adds clipped reference-policy ratio to limit drift, reporting better results than standard SFT and RL on math and code benchmarks.