When Reward Hacking Rebounds: Understanding and Mitigating It with Representation-Level Signals

• URL: http://arxiv.org/abs/2604.01476v1
• Date: Wed, 01 Apr 2026 23:33:08 GMT
• Title: When Reward Hacking Rebounds: Understanding and Mitigating It with Representation-Level Signals
• Abstract summary: Reinforcement learning for LLMs is vulnerable to reward hacking.<n>We study this phenomenon in coding tasks using an environment-manipulation setting.<n>We propose Advantage Modification, which integrates shortcut concept scores into GRPO advantage computation.
• Score: 11.280037154530847
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reinforcement learning for LLMs is vulnerable to reward hacking, where models exploit shortcuts to maximize reward without solving the intended task. We systematically study this phenomenon in coding tasks using an environment-manipulation setting, where models can rewrite evaluator code to trivially pass tests without solving the task, as a controlled testbed. Across both studied models, we identify a reproducible three-phase rebound pattern: models first attempt to rewrite the evaluator but fail, as their rewrites embed test cases their own solutions cannot pass. They then temporarily retreat to legitimate solving. When legitimate reward remains scarce, they rebound into successful hacking with qualitatively different strategies. Using representation engineering, we extract concept directions for shortcut, deception, and evaluation awareness from domain-general contrastive pairs and find that the shortcut direction tracks hacking behavior most closely, making it an effective representational proxy for detection. Motivated by this finding, we propose Advantage Modification, which integrates shortcut concept scores into GRPO advantage computation to penalize hacking rollouts before policy updates. Because the penalty is internalized into the training signal rather than applied only at inference time, Advantage Modification provides more robust suppression of hacking compared with generation-time activation steering.

Related papers

• When Do Intrinsic Rewards Work for Code Reasoning? A Comprehensive Study [77.64957118012423]
  Reinforcement learning with verifiable rewards (RLVR) has driven substantial progress in large language model reasoning.<n>Recent work addresses this by deriving rewards from a model's own signals, such as majority voting or confidence-based scores.<n>We present a systematic empirical study of intrinsic reward methods for code generation.
  arXiv  Detail & Related papers  (2026-06-18T19:15:50Z)
• Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models [59.427161616512855]
  Reward hacking arises when a model improves a proxy reward by exploiting shortcuts rather than solving the intended task.<n>We analyze this drift through dominant singular directions of parameter updates and show that reward-hacking runs exhibit substantially larger directional change than clean runs.
  arXiv  Detail & Related papers  (2026-05-24T17:34:34Z)
• Detecting and Suppressing Reward Hacking with Gradient Fingerprints [46.25008147178368]
  Fingerprint Gradient (GRIFT) is a method for detecting reward hacking using models' internal computations.<n>Across verifiable reasoning benchmarks, GRIFT substantially outperforms strong baselines.<n>Our results highlight a promising direction of leveraging gradient level representations for assessing the quality of CoT reasoning traces.
  arXiv  Detail & Related papers  (2026-04-17T17:01:24Z)
• IR$^3$: Contrastive Inverse Reinforcement Learning for Interpretable Detection and Mitigation of Reward Hacking [67.20568716300272]
  Reinforcement Learning from Human Feedback (RLHF) enables powerful LLM alignment but can introduce reward hacking.<n>We introduce IR3 (Interpretable Reward Reconstruction and Rectification), a framework that reverse-engineers, interprets, and surgically repairs the implicit objectives driving RLHF-tuned models.<n>We show that IR3 achieves 0.89 correlation with ground-truth rewards, identifies hacking features with over 90% precision, and significantly reduces hacking behaviors while maintaining capabilities within 3% of the original model.
  arXiv  Detail & Related papers  (2026-02-23T01:14:53Z)
• Reward Under Attack: Analyzing the Robustness and Hackability of Process Reward Models [68.45272703833209]
  We show that state-of-the-art PRMs are systematically exploitable under adversarial optimization pressure.<n>We introduce a three-tiered diagnostic framework that applies increasing adversarial pressure to quantify these vulnerabilities.<n>We release PRM-BiasBench and a diagnostic toolkit to enable robustness evaluation before deployment.
  arXiv  Detail & Related papers  (2026-02-20T23:38:03Z)
• Adversarial Reward Auditing for Active Detection and Mitigation of Reward Hacking [69.06218054848803]
  We propose Adrial Reward Auditing (ARA), a framework that reconceptualizes reward hacking as a dynamic, competitive game.<n>ARA operates in two stages: first, a Hacker policy discovers reward model vulnerabilities while an Auditor learns to detect exploitation from latent representations.<n>ARA achieves the best alignment-utility tradeoff among all baselines.
  arXiv  Detail & Related papers  (2026-02-02T07:34:57Z)
• Rectifying Shortcut Behaviors in Preference-based Reward Learning [46.09046818725698]
  In reinforcement learning, preference-based reward models play a central role in aligning large language models to human-aligned behavior.<n>Recent studies show that these models are prone to reward hacking and often fail to generalize well due to over-optimization.<n>We introduce a principled yet flexible approach to mitigate shortcut behaviors in preference-based reward learning.
  arXiv  Detail & Related papers  (2025-10-21T20:08:32Z)
• LaSeR: Reinforcement Learning with Last-Token Self-Rewarding [54.72617309922891]
  Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a core paradigm for enhancing the reasoning capabilities of Large Language Models (LLMs)<n>Previous practice requires the LLM to sequentially generate solutions and self-verifications using two separate prompt templates, which significantly reduces efficiency.<n>We propose LaSeR (Reinforcement Learning with Last-Token Self-Rewarding), an algorithm that simply augments the original RLVR loss with a MSE loss.
  arXiv  Detail & Related papers  (2025-10-16T17:55:11Z)
• Learning a Dense Reasoning Reward Model from Expert Demonstration via Inverse Reinforcement Learning [50.20267980386502]
  We learn a dense, token-level reward model for process supervision directly from expert demonstrations.<n>The learned reasoning reward serves two complementary roles: (i) it provides step-level feedback to optimise a reasoning policy during training; and (ii) it functions at inference as a critic to rerank sampled traces under fixed compute budgets.
  arXiv  Detail & Related papers  (2025-10-02T09:55:26Z)
• Reinforcement Learning From Imperfect Corrective Actions And Proxy Rewards [38.056359612828466]
  We propose a novel value-based deep RL algorithm called Iterative learning from Corrective actions and Proxy rewards (ICoPro) We experimentally validate our proposition on a variety of tasks (Atari games and autonomous driving on highway)
  arXiv  Detail & Related papers  (2024-10-08T08:04:09Z)
• A State Augmentation based approach to Reinforcement Learning from Human Preferences [20.13307800821161]
  Preference Based Reinforcement Learning attempts to solve the issue by utilizing binary feedbacks on queried trajectory pairs. We present a state augmentation technique that allows the agent's reward model to be robust.
  arXiv  Detail & Related papers  (2023-02-17T07:10:50Z)
• Backdoor Defense via Suppressing Model Shortcuts [91.30995749139012]
  In this paper, we explore the backdoor mechanism from the angle of the model structure. We demonstrate that the attack success rate (ASR) decreases significantly when reducing the outputs of some key skip connections.
  arXiv  Detail & Related papers  (2022-11-02T15:39:19Z)

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.