Skip to content
← Back to explorer

RM-R1: Reward Modeling as Reasoning

Xiusi Chen, Gaotang Li, Ziqi Wang, Bowen Jin, Cheng Qian, Yu Wang, Hongru Wang, Yu Zhang, Denghui Zhang, Tong Zhang, Hanghang Tong, Heng Ji · May 5, 2025 · Citations: 0

Coding Math Pairwise Preference Rubric Rating
Open arXiv Find Implementation RSS feed Shortlist (0)

Data freshness

Extraction: Fresh

Check recency before relying on this page for active eval decisions. Use stale pages as context and verify against current hub results.

Metadata refreshed

Mar 6, 2026, 7:08 AM

Recent

Extraction refreshed

Mar 14, 2026, 12:16 AM

Fresh

Extraction source

Persisted extraction

Confidence 0.45

Abstract

Reward modeling is essential for aligning large language models with human preferences through reinforcement learning. To provide accurate reward signals, a reward model (RM) should stimulate deep thinking and conduct interpretable reasoning before assigning a score or a judgment. Inspired by recent advances of long chain-of-thought on reasoning-intensive tasks, we hypothesize and validate that integrating reasoning into reward modeling significantly enhances RM's interpretability and performance. We introduce a new class of generative reward models, Reasoning Reward Models (ReasRMs), which formulate reward modeling as a reasoning task. We propose a reasoning-oriented training pipeline and train a family of ReasRMs, RM-R1. RM-R1 features a chain-of-rubrics (CoR) mechanism -- self-generating sample-level chat rubrics or math/code solutions, and evaluating candidate responses against them. The training of RM-R1 consists of two key stages: (1) distillation of high-quality reasoning chains and (2) reinforcement learning with verifiable rewards. Empirically, our models achieve superior performance across three reward model benchmarks on average, outperforming much larger open-weight models (e.g., INF-ORM-Llama3.1-70B) and proprietary ones (e.g., GPT-4o) by up to 4.9%. Beyond final performance, we perform thorough analyses to understand the key ingredients of successful ReasRM training.

Low-signal caution for protocol decisions

Use this page for context, then validate protocol choices against stronger HFEPX references before implementation decisions.

  • Extraction confidence is 0.45 (below strong-reference threshold).
  • No explicit evaluation mode was extracted from available metadata.
  • No benchmark/dataset or metric anchors were extracted.
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

HFEPX Relevance Assessment

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Best use

Background context only

Use if you need

Background context only.

Main weakness

Extraction confidence is 0.45 (below strong-reference threshold).

Trust level

Low

Eval-Fit Score

40/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Detected

Evaluation Signal

Weak / implicit signal

HFEPX Fit

Adjacent candidate

Extraction confidence: Low

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Field Provenance & Confidence

Each key protocol field shows extraction state, confidence band, and data source so you can decide whether to trust it directly or validate from full text.

Human Feedback Types

partial

Pairwise Preference, Rubric Rating

Confidence: Low Source: Persisted extraction evidenced

Directly usable for protocol triage.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Evaluation Modes

missing

None explicit

Confidence: Low Source: Persisted extraction missing

Validate eval design from full paper text.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Quality Controls

missing

Not reported

Confidence: Low Source: Persisted extraction missing

No explicit QC controls found.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Benchmarks / Datasets

missing

Not extracted

Confidence: Low Source: Persisted extraction missing

No benchmark anchors detected.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Reported Metrics

missing

Not extracted

Confidence: Low Source: Persisted extraction missing

No metric anchors detected.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Rater Population

missing

Unknown

Confidence: Low Source: Persisted extraction missing

Rater source not explicitly reported.

Evidence snippet: Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Human Data Lens

  • Uses human feedback: Yes
  • Feedback types: Pairwise Preference, Rubric Rating
  • Rater population: Unknown
  • Unit of annotation: Multi Dim Rubric
  • Expertise required: Math, Coding
  • Extraction source: Persisted extraction

Evaluation Lens

  • Evaluation modes:
  • Agentic eval: None
  • Quality controls: Not reported
  • Confidence: 0.45
  • Flags: ambiguous

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

No metric terms were extracted from the available abstract.

Research Brief

Deterministic synthesis

We introduce a new class of generative reward models, Reasoning Reward Models (ReasRMs), which formulate reward modeling as a reasoning task. HFEPX signals include Pairwise Preference, Rubric Rating with confidence 0.45. Updated from current HFEPX corpus.

Generated Mar 14, 2026, 12:16 AM · Grounded in abstract + metadata only

Key Takeaways

  • We introduce a new class of generative reward models, Reasoning Reward Models (ReasRMs), which formulate reward modeling as a reasoning task.
  • We propose a reasoning-oriented training pipeline and train a family of ReasRMs, RM-R1.
  • Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.

Researcher Actions

  • Compare its human-feedback setup against pairwise and rubric hubs.
  • Identify benchmark choices from full text before operationalizing conclusions.
  • Verify metric definitions before comparing against your eval pipeline.

Caveats

  • Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
  • Extraction confidence is probabilistic and should be validated for critical decisions.

Research Summary

Contribution Summary

  • We introduce a new class of generative reward models, Reasoning Reward Models (ReasRMs), which formulate reward modeling as a reasoning task.
  • We propose a reasoning-oriented training pipeline and train a family of ReasRMs, RM-R1.
  • Empirically, our models achieve superior performance across three reward model benchmarks on average, outperforming much larger open-weight models (e.g., INF-ORM-Llama3.1-70B) and proprietary ones (e.g., GPT-4o) by up to 4.9%.

Why It Matters For Eval

  • Reward modeling is essential for aligning large language models with human preferences through reinforcement learning.
  • Empirically, our models achieve superior performance across three reward model benchmarks on average, outperforming much larger open-weight models (e.g., INF-ORM-Llama3.1-70B) and proprietary ones (e.g., GPT-4o) by up to 4.9%.

Researcher Checklist

  • Pass: Human feedback protocol is explicit

    Detected: Pairwise Preference, Rubric Rating

  • Gap: Evaluation mode is explicit

    No clear evaluation mode extracted.

  • Gap: Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Gap: Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Gap: Metric reporting is present

    No metric terms extracted.

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote