# Review Rarely Beats the Stronger Model: Separating Verification from Capability Transfer in LLM Pipelines

arXiv: https://arxiv.org/abs/2610.15897 · HTML: https://mlsanity.com/paper/2610.15897/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 32 of its 58 references that are in the map.
- First posted on arXiv 11 Oct 2026.

## Builds on (most cited first, 30 of 32)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [Large Language Models are not Fair Evaluators](https://mlsanity.com/paper/2305.17926/) (2023)
- [Large Language Models Cannot Self-Correct Reasoning Yet](https://mlsanity.com/paper/2310.01798/) (2023)
- [FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance](https://mlsanity.com/paper/2305.05176/) (2023)
- [LLM Evaluators Recognize and Favor Their Own Generations](https://mlsanity.com/paper/2404.13076/) (2024)
- [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024)
- [LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion](https://mlsanity.com/paper/2306.02561/) (2023)
- [Hybrid LLM: Cost-Efficient and Quality-Aware Query Routing](https://mlsanity.com/paper/2404.14618/) (2024)
- [Generating Sequences by Learning to Self-Correct](https://mlsanity.com/paper/2211.00053/) (2022)
- [When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMs](https://mlsanity.com/paper/2406.01297/) (2024)
- [ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning](https://mlsanity.com/paper/2002.04326/) (2020)
- [Power Hungry Processing: Watts Driving the Cost of AI Deployment?](https://mlsanity.com/paper/2311.16863/) (2023)
- [Large Language Model Routing with Benchmark Datasets](https://mlsanity.com/paper/2309.15789/) (2023)
- [Multi-Agent Verification: Scaling Test-Time Compute with Multiple Verifiers](https://mlsanity.com/paper/2502.20379/) (2025)
- [Seed-Coder: Let the Code Model Curate Data for Itself](https://mlsanity.com/paper/2506.03524/) (2025)
- [When To Solve, When To Verify: Compute-Optimal Problem Solving and Generative Verification for LLM Reasoning](https://mlsanity.com/paper/2504.01005/) (2025)
- [Conformity in Large Language Models](https://mlsanity.com/paper/2410.12428/) (2024)
- [JEV-as-a-Judge: Accept When Confident, Escalate When Unsure](https://mlsanity.com/paper/2609.26550/) (2026)
- [When Identity Skews Debate: Anonymization for Bias-Reduced Multi-Agent Reasoning](https://mlsanity.com/paper/2510.07517/) (2025)
- [When Does Verification Pay Off? A Closer Look at LLMs as Solution Verifiers](https://mlsanity.com/paper/2512.02304/) (2025)
- [More Convincing, Not More Correct: Self-Play Reward Hacking of Reference-Free LLM Judges](https://mlsanity.com/paper/2607.05904/) (2026)
- [Easier to Mislead Than to Correct: Harmful and Beneficial Revision in LLM Conformity](https://mlsanity.com/paper/2606.01637/) (2026)
- [Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems](https://mlsanity.com/paper/2604.02668/) (2026)
- [Cross-Model LLM Code Review: Should you use Claude to review Codex or vice versa?](https://mlsanity.com/paper/2607.21656/) (2026)
- [When to Screen, When to Bypass: LLM-Judges in Resource-Scarce AI-Human Workflow](https://mlsanity.com/paper/2603.13870/) (2026)

## Abstract (arXiv)

Qwen3-8B reviewing Llama-3.1-8B in math adds 33.2 points, yet Qwen3-8B alone is 6.4 points more accurate than the result. Reviewing another model's answer is usually judged by its gain over the generator. We show that this confuses verification, which adds accuracy beyond either model alone, with capability transfer, where a stronger reviewer supplies answers it could have given on its own. Measured against the stronger model alone, across 48 pairs of 12 open-weight models (3B to 72B) in math, knowledge, code and reasoning, review beats it in only 9 of 96 review directions, and does worse the larger the capability gap. The reason is anchoring: the reviewer mostly follows the generator, discarding its own correct answers. Review costs 2.6 to 2.8 times as much as the stronger model alone, while a selector using the same two calls is 2.7 to 4.2 points more accurate than review and cheaper. Studies of verification should report the stronger model alone, not only the gain over the generator.