Pith. sign in

American invitational mathematics examination - aime

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

1 Pith paper citing it

fields

cs.LG 1

years

2025 1

verdicts

CONDITIONAL 1

representative citing papers

On-Policy RL with Optimal Reward Baseline

cs.LG · 2025-05-29 · conditional · novelty 4.0

A length-weighted reward baseline plus exact on-policy updates improves stability and diversity in LLM reinforcement learning for math, with small performance gains over GRPO.

citing papers explorer

Showing 1 of 1 citing paper.

  • On-Policy RL with Optimal Reward Baseline cs.LG · 2025-05-29 · conditional · none · ref 8

    A length-weighted reward baseline plus exact on-policy updates improves stability and diversity in LLM reinforcement learning for math, with small performance gains over GRPO.