# Post-hoc Reward Calibration: A Case Study on Length Bias

arXiv: https://arxiv.org/abs/2409.17407 · HTML: https://mlsanity.com/paper/2409.17407/

## Key facts

- Cited by 29 papers in the map (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 62 of its 72 references that are in the map.
- First posted on arXiv 25 Sep 2024.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 30 of 62)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them](https://mlsanity.com/paper/2210.09261/) (2022)
- [Instruction-Following Evaluation for Large Language Models](https://mlsanity.com/paper/2311.07911/) (2023)
- [Scaling Laws for Reward Model Overoptimization](https://mlsanity.com/paper/2210.10760/) (2022)
- [SimPO: Simple Preference Optimization with a Reference-Free Reward](https://mlsanity.com/paper/2405.14734/) (2024)
- [Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluators](https://mlsanity.com/paper/2404.04475/) (2024)
- [AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback](https://mlsanity.com/paper/2305.14387/) (2023)
- [Aligning AI With Shared Human Values](https://mlsanity.com/paper/2008.02275/) (2020)
- [Aligning Large Multimodal Models with Factually Augmented RLHF](https://mlsanity.com/paper/2309.14525/) (2023)
- [RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment](https://mlsanity.com/paper/2304.06767/) (2023)
- [Safe RLHF: Safe Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2310.12773/) (2023)
- [Prometheus: Inducing Fine-grained Evaluation Capability in Language Models](https://mlsanity.com/paper/2310.08491/) (2023)
- [From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline](https://mlsanity.com/paper/2406.11939/) (2024)
- [RewardBench: Evaluating Reward Models for Language Modeling](https://mlsanity.com/paper/2403.13787/) (2024)
- [UltraFeedback: Boosting Language Models with Scaled AI Feedback](https://mlsanity.com/paper/2310.01377/) (2023)
- [Fine-Grained Human Feedback Gives Better Rewards for Language Model Training](https://mlsanity.com/paper/2306.01693/) (2023)
- [Interpretable Preferences via Multi-Objective Reward Modeling and Mixture-of-Experts](https://mlsanity.com/paper/2406.12845/) (2024)
- [A Long Way to Go: Investigating Length Correlations in RLHF](https://mlsanity.com/paper/2310.03716/) (2023)
- [RLHF Workflow: From Reward Modeling to Online RLHF](https://mlsanity.com/paper/2405.07863/) (2024)
- [A Survey of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2312.14925/) (2023)
- [Reward Design with Language Models](https://mlsanity.com/paper/2303.00001/) (2023)

## Built on by (most cited first, 29 of 29)

- [Reward Shaping to Mitigate Reward Hacking in RLHF](https://mlsanity.com/paper/2502.18770/) (2025)
- [Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling](https://mlsanity.com/paper/2507.01679/) (2025)
- [Beyond Reward Hacking: Causal Rewards for Large Language Model Alignment](https://mlsanity.com/paper/2501.09620/) (2025)
- [Beyond Consensus: Mitigating the Agreeableness Bias in LLM Judge Evaluations](https://mlsanity.com/paper/2510.11822/) (2025)
- [Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanations](https://mlsanity.com/paper/2504.05294/) (2025)
- [CHARM: Calibrating Reward Models With Chatbot Arena Scores](https://mlsanity.com/paper/2504.10045/) (2025)
- [Bias Fitting to Mitigate Length Bias of Reward Model in RLHF](https://mlsanity.com/paper/2505.12843/) (2025)
- [Eliminating Inductive Bias in Reward Models with Information-Theoretic Guidance](https://mlsanity.com/paper/2512.23461/) (2025)
- [Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation](https://mlsanity.com/paper/2602.09305/) (2026)
- [Tackling Length Inflation Without Trade-offs: Group Relative Reward Rescaling for Reinforcement Learning](https://mlsanity.com/paper/2603.10535/) (2026)
- [Mitigating Length Bias in RLHF through a Causal Lens](https://mlsanity.com/paper/2511.12573/) (2025)
- [CoLD: Counterfactually-Guided Length Debiasing for Process Reward Models in Mathematical Reasoning](https://mlsanity.com/paper/2507.15698/) (2025)
- [Energy-Based Reward Models for Robust Language Model Alignment](https://mlsanity.com/paper/2504.13134/) (2025)
- [TuneJury: An Open Metric for Improving Music Generation Preference Alignment](https://mlsanity.com/paper/2606.17006/) (2026)
- [A Judge Should Know What Changed:Construct Validity for LLM-as-a-Judge Evaluation](https://mlsanity.com/paper/2608.24419/) (2026)
- [Multimodal Graph Representation Learning for Robust Surgical Workflow Recognition with Adversarial Feature Disentanglement](https://mlsanity.com/paper/2505.01766/) (2025)
- [One Bias After Another: Mechanistic Reward Shaping and Persistent Biases in Language Reward Models](https://mlsanity.com/paper/2603.03291/) (2026)
- [SteerRM: Debiasing Reward Models via Sparse Autoencoders](https://mlsanity.com/paper/2603.12795/) (2026)
- [Data Selection for LLM Alignment Using Fine-Grained Preferences](https://mlsanity.com/paper/2508.07638/) (2025)
- [When Distance Distracts: Representation Distance Bias in BT-Loss for Reward Models](https://mlsanity.com/paper/2512.06343/) (2025)
- [Balanced Thinking: Improving Chain of Thought Training in Vision Language Models](https://mlsanity.com/paper/2603.18656/) (2026)
- [Context Training with Active Information Seeking](https://mlsanity.com/paper/2605.13050/) (2026)
- [Debiasing Reward Models via Causally Motivated Inference-Time Intervention](https://mlsanity.com/paper/2604.27495/) (2026)
- [Configurable Reward Model for Balanced Safety Alignment](https://mlsanity.com/paper/2605.30487/) (2026)
- [PAFO: Pareto Fairness Optimization for Personalized Reward Modeling](https://mlsanity.com/paper/2606.07988/) (2026)
- [Brevity is the Soul of Inference Efficiency: Inducing Concision in VLMs via Data Curation](https://mlsanity.com/paper/2606.25432/) (2026)
- [BV-Blend: Uncertainty-Weighted Historical Baselines for Stable Critic-Free RL with Verifiable Rewards](https://mlsanity.com/paper/2606.28707/) (2026)
- [What do Reward Models Memorize?](https://mlsanity.com/paper/2607.24484/) (2026)
- [Revisiting On-policy Adversarial Black-Box Distillation: Calibrating Groupwise Reward Geometry for Effective Advantage Construction](https://mlsanity.com/paper/2609.39757/) (2026)

## Abstract (arXiv)

Reinforcement Learning from Human Feedback aligns the outputs of Large Language Models with human values and preferences. Central to this process is the reward model (RM), which translates human feedback into training signals for optimising LLM behaviour. However, RMs can develop biases by exploiting spurious correlations in their training data, such as favouring outputs based on length or style rather than true quality. These biases can lead to incorrect output rankings, sub-optimal model evaluations, and the amplification of undesirable behaviours in LLMs alignment. This paper addresses the challenge of correcting such biases without additional data and training, introducing the concept of Post-hoc Reward Calibration. We first propose an intuitive approach to estimate the bias term and, thus, remove it to approximate the underlying true reward. We then extend the approach to a more general and robust form with the Locally Weighted Regression. Focusing on the prevalent length bias, we validate our proposed approaches across three experimental settings, demonstrating consistent improvements: (1) a 3.11 average performance gain across 33 reward models on the RewardBench dataset; (2) enhanced alignment of RM rankings with GPT-4 evaluations and human preferences based on the AlpacaEval benchmark; and (3) improved Length-Controlled win rate of the RLHF process in multiple LLM--RM combinations. Our method is computationally efficient and generalisable to other types of bias and RMs, offering a scalable and robust solution for mitigating biases in LLM alignment. Our code and results are available at https://github.com/ZeroYuHuang/Reward-Calibration.