SPARK: Stepwise Process-Aware Rewards for Reference-Free Reinforcement Learning

• URL: http://arxiv.org/abs/2512.03244v1
• Date: Tue, 02 Dec 2025 21:30:47 GMT
• Title: SPARK: Stepwise Process-Aware Rewards for Reference-Free Reinforcement Learning
• Authors: Salman Rahman, Sruthi Gorantla, Arpit Gupta, Swastik Roy, Nanyun Peng, Yang Liu,
• Abstract summary: Process reward models (PRMs) that provide dense, step-level feedback have shown promise for reinforcement learning.<n>We propose SPARK: a three-stage framework where in the first stage a generator model produces diverse solutions and a verifier model evaluates them.<n>We show that aggregating multiple independent verifications at the step level produces training data for process reward models that surpass ground-truth outcome supervision.
• Score: 39.1720897614261
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Process reward models (PRMs) that provide dense, step-level feedback have shown promise for reinforcement learning, yet their adoption remains limited by the need for expensive step-level annotations or ground truth references. We propose SPARK: a three-stage framework where in the first stage a generator model produces diverse solutions and a verifier model evaluates them using parallel scaling (self-consistency) and sequential scaling (meta-critique). In the second stage, we use these verification outputs as synthetic training data to fine-tune generative process reward models, which subsequently serve as reward signals during training. We show that aggregating multiple independent verifications at the step level produces training data for process reward models that surpass ground-truth outcome supervision, achieving 67.5 F1 on ProcessBench (a benchmark for identifying erroneous steps in mathematical reasoning) compared to 66.4 for reference-guided training and 61.9 for GPT-4o. In the final stage, we apply our generative PRM with chain-of-thought verification (PRM-CoT) as the reward model in RL experiments on mathematical reasoning, and introduce format constraints to prevent reward hacking. Using Qwen2.5-Math-7B, we achieve 47.4% average accuracy across six mathematical reasoning benchmarks, outperforming ground-truth-based RLVR (43.9%). Our work enables reference-free RL training that exceeds ground-truth methods, opening new possibilities for domains lacking verifiable answers or accessible ground truth.

Related papers

• A Mathematical Framework for Custom Reward Functions in Job Application Evaluation using Reinforcement Learning [0.8709142317087694]
  This article describes a new two-step process to design a more refined resume evaluation model based on a small language model.<n>The model is finetuned using GRPO on a custom reward function.<n>Our resulting GRPO-polished model demonstrates significant real-world efficacy, achieving a final accuracy of 91% on unseen test data.
  arXiv  Detail & Related papers  (2025-11-20T06:06:30Z)
• Curing Miracle Steps in LLM Mathematical Reasoning with Rubric Rewards [40.905635870672945]
  Large language models for mathematical reasoning are typically trained with outcome-based rewards, which credit only the final answer.<n>In our experiments, we observe that this paradigm is highly susceptible to reward hacking, leading to a substantial overestimation of a model's reasoning ability.<n>This is evidenced by a high incidence of false positives - solutions that reach the correct final answer through an unsound reasoning process.
  arXiv  Detail & Related papers  (2025-10-09T04:30:45Z)
• ReasonFlux-PRM: Trajectory-Aware PRMs for Long Chain-of-Thought Reasoning in LLMs [75.72672339168092]
  We introduce ReasonFlux-PRM, a novel trajectory-aware PRM to evaluate trajectory-response type of reasoning traces.<n>ReasonFlux-PRM incorporates both step-level and trajectory-level supervision, enabling fine-grained reward assignment aligned with structured chain-of-thought data.<n>Our derived ReasonFlux-PRM-7B yields consistent performance improvements, achieving average gains of 12.1% in supervised fine-tuning, 4.5% in reinforcement learning, and 6.3% in test-time scaling.
  arXiv  Detail & Related papers  (2025-06-23T17:59:02Z)
• Ring-lite: Scalable Reasoning via C3PO-Stabilized Reinforcement Learning for LLMs [51.21041884010009]
  Ring-lite is a Mixture-of-Experts (MoE)-based large language model optimized via reinforcement learning (RL)<n>Our approach matches the performance of state-of-the-art (SOTA) small-scale reasoning models on challenging benchmarks.
  arXiv  Detail & Related papers  (2025-06-17T17:12:34Z)
• Discriminative Policy Optimization for Token-Level Reward Models [55.98642069903191]
  Process reward models (PRMs) provide more nuanced supervision compared to outcome reward models (ORMs)<n>Q-RM explicitly learns token-level Q-functions from preference data without relying on fine-grained annotations.<n>Reinforcement learning with Q-RM significantly enhances training efficiency, achieving convergence 12 times faster than ORM on GSM8K and 11 times faster than step-level PRM on MATH.
  arXiv  Detail & Related papers  (2025-05-29T11:40:34Z)
• Stop Summation: Min-Form Credit Assignment Is All Process Reward Model Needs for Reasoning [36.320502608626896]
  Process reward models (PRMs) have proven effective for test-time scaling of Large Language Models (LLMs) on challenging reasoning tasks.<n>However, reward hacking issues with PRMs limit their successful application in reinforcement fine-tuning.<n>In this paper, we identify the main cause of PRM-induced reward hacking: the canonical summation-form credit assignment in reinforcement learning.
  arXiv  Detail & Related papers  (2025-04-21T17:59:02Z)
• Exploring the Limit of Outcome Reward for Learning Mathematical Reasoning [65.2421542320293]
  Reasoning abilities are crucial components of general intelligence.<n>Recent advances by proprietary companies, such as o-series models of OpenAI, have made remarkable progress on reasoning tasks.<n>This paper proposes a new RL framework, termed OREAL, to pursue the performance limit that can be achieved through textbfOutcome textbfREwtextbfArd-based reinforcement textbfLearning for mathematical reasoning tasks.
  arXiv  Detail & Related papers  (2025-02-10T18:57:29Z)
• ACECODER: Acing Coder RL via Automated Test-Case Synthesis [36.740393665032954]
  We design a pipeline that generates extensive (question, test-cases) pairs from existing code data.<n>We construct preference pairs based on pass rates over sampled programs to train reward models with Bradley-Terry loss.<n>We show that our RL training can improve model on HumanEval-plus by over 25% and MBPP-plus by 6% for merely 80 optimization steps.
  arXiv  Detail & Related papers  (2025-02-03T18:46:04Z)
• Self-Taught Evaluators [77.92610887220594]
  We present an approach that aims to im-proves without human annotations, using synthetic training data only. Our Self-Taught Evaluator can improve a strong LLM from 75.4 to 88.3 on RewardBench.
  arXiv  Detail & Related papers  (2024-08-05T17:57:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.