# From Rebound to Remedy: Understanding and Mitigating Reward Hacking via Representation Engineering Source: [From Rebound to Remedy: Understanding and Mitigating Reward Hacking via Representation Engineering]() Collection index: [llms.txt]() Data updated: 2026-08-12T05:00:13.408Z Authors: Rui Wu, Ruixiang Tang Published: 2026-04-01T23:33:08.000Z ## 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. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Code Generation on LeetCode without write access](): Code Generation; LeetCode without write access; Pass@1: 15. - [Reward Hacking on LeetCode with write access (test)](): Reward Hacking; LeetCode with write access (test); Hack Rate: 1.2. ## Original sources and implementation [Original paper]() [PDF]()