{"ID":2890866,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.18305","arxiv_id":"2507.18305","title":"BadReasoner: Planting Tunable Overthinking Backdoors into Large Reasoning Models for Fun or Profit","abstract":"Large reasoning models (LRMs) have emerged as a significant advancement in artificial intelligence, representing a specialized class of large language models (LLMs) designed to tackle complex reasoning tasks. The defining characteristic of LRMs lies in their extensive chain-of-thought (CoT) reasoning capabilities. In this paper, we identify a previously unexplored attack vector against LRMs, which we term \"overthinking backdoors\". We advance this concept by proposing a novel tunable backdoor, which moves beyond simple on/off attacks to one where an attacker can precisely control the extent of the model's reasoning verbosity. Our attack is implemented through a novel data poisoning methodology. It pairs a tunable trigger-where the number of repetitions signals the desired intensity-with a correspondingly verbose CoT response. These responses are programmatically generated by instructing a teacher LLM to inject a controlled number of redundant refinement steps into a correct reasoning process. The approach preserves output correctness, which ensures stealth and establishes the attack as a pure resource-consumption vector. Extensive empirical results on various LRMs demonstrate that our method can reliably trigger a controllable, multi-fold increase in the length of the reasoning process, without degrading the final answer's correctness. Our source code is available at https://github.com/FZaKK/BadReasoner.","short_abstract":"Large reasoning models (LRMs) have emerged as a significant advancement in artificial intelligence, representing a specialized class of large language models (LLMs) designed to tackle complex reasoning tasks. The defining characteristic of LRMs lies in their extensive chain-of-thought (CoT) reasoning capabilities. In t...","url_abs":"https://arxiv.org/abs/2507.18305","url_pdf":"https://arxiv.org/pdf/2507.18305v1","authors":"[\"Biao Yi\",\"Zekun Fei\",\"Jianing Geng\",\"Tong Li\",\"Lihai Nie\",\"Zheli Liu\",\"Yiming Li\"]","published":"2025-07-24T11:24:35Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":611816,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2890866,"paper_url":"https://arxiv.org/abs/2507.18305","paper_title":"BadReasoner: Planting Tunable Overthinking Backdoors into Large Reasoning Models for Fun or Profit","repo_url":"https://github.com/FZaKK/BadReasoner","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}