Backdoors in RLVR: Jailbreak Backdoors in LLMs From Verifiable Reward
This work reveals a new security threat for LLMs trained with RLVR, a paradigm used to improve reasoning, by demonstrating that backdoors can be injected via data poisoning alone.
The paper identifies a vulnerability in Reinforcement Learning with Verifiable Rewards (RLVR) for LLMs, showing that a backdoor can be implanted using less than 2% poisoned training data without modifying the reward verifier. The attack degrades safety performance by an average of 73% across jailbreak benchmarks while maintaining benign task performance.
Reinforcement Learning with Verifiable Rewards (RLVR) is an emerging paradigm that significantly boosts a Large Language Model's (LLM's) reasoning abilities on complex logical tasks, such as mathematics and programming. However, we identify, for the first time, a latent vulnerability to backdoor attacks within the RLVR framework. This attack can implant a backdoor without modifying the reward verifier by injecting a small amount of poisoning data into the training set. Specifically, we propose a novel trigger mechanism designated as the \ourapproach (ACB). The attack exploits the RLVR training loop by assigning substantial positive rewards for harmful responses and negative rewards for refusals. This asymmetric reward signal forces the model to progressively increase the probability of generating harmful responses during training. Our findings demonstrate that the RLVR backdoor attack is characterized by both high efficiency and strong generalization capabilities. Utilizing less than 2\% poisoned data in train set, the backdoor can be successfully implanted across various model scales without degrading performance on benign tasks. Evaluations across multiple jailbreak benchmarks indicate that activating the trigger degrades safety performance by an average of 73\%. Furthermore, the attack generalizes effectively to a wide range of jailbreak methods and unsafe behaviors. Code is available at https://github.com/yuki-younai/Backdoor_in_RLVR.