Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanations

• URL: http://arxiv.org/abs/2504.05294v1
• Date: Mon, 07 Apr 2025 17:49:23 GMT
• Title: Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanations
• Authors: Pedro Ferreira, Wilker Aziz, Ivan Titov,
• Abstract summary: Chain-of-thought explanations are widely used to inspect the decision process of large language models.<n>We show that preference optimization can inadvertently reduce the faithfulness of these explanations.
• Score: 30.68740512996253
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Chain-of-thought explanations are widely used to inspect the decision process of large language models (LLMs) and to evaluate the trustworthiness of model outputs, making them important for effective collaboration between LLMs and humans. We demonstrate that preference optimization - a key step in the alignment phase - can inadvertently reduce the faithfulness of these explanations. This occurs because the reward model (RM), which guides alignment, is tasked with optimizing both the expected quality of the response and the appropriateness of the explanations (e.g., minimizing bias or adhering to safety standards), creating potential conflicts. The RM lacks a mechanism to assess the consistency between the model's internal decision process and the generated explanation. Consequently, the LLM may engage in "reward hacking" by producing a final response that scores highly while giving an explanation tailored to maximize reward rather than accurately reflecting its reasoning. To address this issue, we propose enriching the RM's input with a causal attribution of the prediction, allowing the RM to detect discrepancies between the generated self-explanation and the model's decision process. In controlled settings, we show that this approach reduces the tendency of the LLM to generate misleading explanations.

Related papers

• Supervised Optimism Correction: Be Confident When LLMs Are Sure [91.7459076316849]
  We establish a novel theoretical connection between supervised fine-tuning and offline reinforcement learning. We show that the widely used beam search method suffers from unacceptable over-optimism. We propose Supervised Optimism Correction, which introduces a simple yet effective auxiliary loss for token-level $Q$-value estimations.
  arXiv  Detail & Related papers  (2025-04-10T07:50:03Z)
• Reward Models Identify Consistency, Not Causality [54.987590763737145]
  State-of-the-art reward models prioritize structural consistency over causal correctness.<n>Removing the problem statement has minimal impact on reward scores.<n> altering numerical values or disrupting the reasoning flow significantly affects RM outputs.
  arXiv  Detail & Related papers  (2025-02-20T14:57:14Z)
• Beyond Reward Hacking: Causal Rewards for Large Language Model Alignment [30.605500809158986]
  We propose a novel causal reward modeling approach that integrates causal inference to mitigate spurious correlations. Our approach mitigates various types of spurious correlations effectively, resulting in more reliable and fair alignment of LLMs with human preferences.
  arXiv  Detail & Related papers  (2025-01-16T16:00:37Z)
• The Lessons of Developing Process Reward Models in Mathematical Reasoning [62.165534879284735]
  Process Reward Models (PRMs) aim to identify and mitigate intermediate errors in the reasoning processes. We develop a consensus filtering mechanism that effectively integrates Monte Carlo (MC) estimation with Large Language Models (LLMs) We release a new state-of-the-art PRM that outperforms existing open-source alternatives.
  arXiv  Detail & Related papers  (2025-01-13T13:10:16Z)
• Reinforcing Thinking through Reasoning-Enhanced Reward Models [6.636512424910708]
  Large Language Models (LLMs) exhibit great potential in complex multi-step reasoning through inference-time thinking.<n>LLMs struggle with deciding when to stop thinking due to limited self-awareness about their knowledge boundaries.<n>This work addresses these challenges by distilling the LLM's own reasoning processes into synthetic behavioral data.
  arXiv  Detail & Related papers  (2024-12-31T04:50:15Z)
• Cycles of Thought: Measuring LLM Confidence through Stable Explanations [53.15438489398938]
  Large language models (LLMs) can reach and even surpass human-level accuracy on a variety of benchmarks, but their overconfidence in incorrect responses is still a well-documented failure mode. We propose a framework for measuring an LLM's uncertainty with respect to the distribution of generated explanations for an answer.
  arXiv  Detail & Related papers  (2024-06-05T16:35:30Z)
• Calibrating Reasoning in Language Models with Internal Consistency [18.24350001344488]
  Large language models (LLMs) have demonstrated impressive capabilities in various reasoning tasks. LLMs often generate text with obvious mistakes and contradictions. In this work, we investigate reasoning in LLMs through the lens of internal representations.
  arXiv  Detail & Related papers  (2024-05-29T02:44:12Z)
• Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoning [38.60086807496399]
  Large language models (LLMs) have been shown to perform better when asked to reason step-by-step before answering a question. It is unclear to what degree the model's final answer is faithful to the stated reasoning steps. We introduce FRODO, a framework to tailor small-sized LMs to generate correct reasoning steps and robustly reason over these steps.
  arXiv  Detail & Related papers  (2024-02-21T17:23:59Z)
• FaithLM: Towards Faithful Explanations for Large Language Models [67.29893340289779]
  Large Language Models (LLMs) have become proficient in addressing complex tasks by leveraging their internal knowledge and reasoning capabilities. The black-box nature of these models complicates the task of explaining their decision-making processes. We introduce FaithLM to explain the decision of LLMs with natural language (NL) explanations.
  arXiv  Detail & Related papers  (2024-02-07T09:09:14Z)
• Question Decomposition Improves the Faithfulness of Model-Generated Reasoning [23.34325378824462]
  Large language models (LLMs) are difficult to verify the correctness and safety of their behavior. One approach is to prompt LLMs to externalize their reasoning, by having them generate step-by-step reasoning as they answer a question. This approach relies on the stated reasoning faithfully reflecting the model's actual reasoning, which is not always the case. Decomposition-based methods achieve strong performance on question-answering tasks, sometimes approaching that of CoT.
  arXiv  Detail & Related papers  (2023-07-17T00:54:10Z)
• Robustness and Accuracy Could Be Reconcilable by (Proper) Definition [109.62614226793833]
  The trade-off between robustness and accuracy has been widely studied in the adversarial literature. We find that it may stem from the improperly defined robust error, which imposes an inductive bias of local invariance. By definition, SCORE facilitates the reconciliation between robustness and accuracy, while still handling the worst-case uncertainty.
  arXiv  Detail & Related papers  (2022-02-21T10:36:09Z)

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.