Beyond Reward Hacking: Causal Rewards for Large Language Model Alignment

Recent advances in large language models (LLMs) have demonstrated significant progress in performing complex tasks. While Reinforcement Learning from Human Feedback (RLHF) has been effective in aligning LLMs with human preferences, it is susceptible to spurious correlations in reward modeling. Consequently, it often introduces biases—such as length bias, sycophancy, conceptual bias, and discrimination—that hinder the model’s ability to capture true causal relationships. To address this, we propose a novel causal reward modeling approach that integrates causal inference to mitigate these spurious correlations. Our method enforces counterfactual invariance, ensuring reward predictions remain consistent when irrelevant variables are altered. Through experiments on both synthetic and real-world datasets, we show that our approach mitigates various types of spurious

Recent advancements in large language models (LLMs) have demonstrated remarkable capabilities in generating coherent, contextually appropriate responses across a wide range of tasks (Brown et al., 2020). A key approach to further refine these models is Reinforcement Learning from Human Feedback (RLHF), which leverages human evaluations to guide the training process and align model outputs more closely with human preferences (Stiennon et al., 2020; Ouyang et al., 2022; Bai et al., 2022; Wang et al., 2024). RLHF typically involves training a reward model to capture human preferences, which is then used to fine-tune LLMs via reinforcement learning (RL) (Schulman et al., 2017; Chen et al., 2024b,f).

Despite the success of RLHF, reward modeling is inherently prone to spurious correlations, which are associations in the training data that do not reflect true causal relationships (Veitch et al., 2021), and can lead to unintended biases and induce reward hacking (McMilin, 2022). Reward hacking occurs when RL agents exploit flaws or ambiguities in the reward function to maximize rewards without genuinely improving alignment with desired behaviors or completing designed tasks (Amodei et al., 2016; Weng, 2024). Consequently, this leads to misaligned models that exhibit biases such as favoring longer outputs (length bias) (Zheng et al., 2023), agreeing with user’s incorrect assertions (sycophancy bias) (Perez et al., 2022), developing unintended shortcuts when making predictions (concept bias) (Zhou et al., 2023), and implicitly developing discrimination over certain demographic groups (discrimination bias) (Tamkin et al., 2023; Chen et al., 2024c). These biases, rooted in spurious correlations and reward hacking rather than true causal relationships, undermine the reliability and trustworthiness of LLMs, posing significant challenges for their safe and responsible deployment in real-world applications (Anwar et al., 2024; Qi et al., 2024).

To address this challenge, we propose a novel approach in this work that integrates causality into reward modeling to mitigate the impact of spurious correlations and prevent reward hacking in RLHF. By leveraging causal inference techniques, we develop a causalrewardmodel(CRM) that is robust to these spurious correlations and captures the true causal relationship of responses on human preferences. Central to our method is the concept of counterfactual invariance, which ensures that the reward model’s predictions remain consistent under interventions on irrelevant aspects of the input, thereby reducing the irreducible error caused by spurious correlations (Veitch et al., 2021).