pith. sign in

In this section, you must clearly list each degradation method you used, and for each one, pinpoint exactly how, where, and why you altered the original reasoning

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

1 Pith paper citing it

fields

cs.SE 1

years

2025 1

verdicts

UNVERDICTED 1

representative citing papers

ReCode: Reinforcing Code Generation with Reasoning-Process Rewards

cs.SE · 2025-08-07 · unverdicted · novelty 6.0

ReCode is a new RL framework combining contrastive reasoning-process reward learning with consistency-gated GRPO to improve code generation, yielding a 16.1% gain for a 7B model to match GPT-4-Turbo levels on benchmarks.

citing papers explorer

Showing 1 of 1 citing paper.

  • ReCode: Reinforcing Code Generation with Reasoning-Process Rewards cs.SE · 2025-08-07 · unverdicted · none · ref 15

    ReCode is a new RL framework combining contrastive reasoning-process reward learning with consistency-gated GRPO to improve code generation, yielding a 16.1% gain for a 7B model to match GPT-4-Turbo levels on benchmarks.