The paper introduces OverThink, a slowdown attack that forces reasoning language models (RLMs) to produce many more reasoning tokens while still giving correct answers. By injecting decoy reasoning problems—such as Markov decision processes, language translation, or graphic comprehension—into the model’s context, attackers can dramatically increase token generation (up to 46× on SQuAD and 17× on coding agents). The study evaluates the attack on both proprietary and open-source RLMs across multiple datasets, explores multimodal and coding‑agent variants, and tests several defenses, concluding that defending against OverThink is challenging and that newer RLMs are even more vulnerable due to higher per‑token costs and increased reasoning token usage.
By Abhinav Kumar, Jaechul Roh, Ali Naseh, Marzena Karpinska, Mohit Iyyer, Amir Houmansadr, Eugene Bagdasarian
arXiv:2606. 02835v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) improve performance by generating explicit intermediate reasoning traces through increased test-time compute, yet the assumption that longer reasoning is consistently beneficial remains under-examined.
By Simone Caldarella, Davide Talon, Rahaf Aljundi, Elisa Ricci, Massimiliano Mancini
arXiv:2602. 14095v2 Announce Type: replace Abstract: Monitoring chain-of-thought (CoT) reasoning is a foundational safety technique for large language model agents; however, this oversight is compromised if models learn to conceal their reasoning.
By Artem Karpov
arXiv:2606. 07410v1 Announce Type: cross Abstract: The emergence of "Aha moments" in large language models, particularly DeepSeek-R1-0120, has raised the question of whether these systems genuinely reason or merely imitate the appearance of reasoning.
By Yuxiang Chen, Jun Wang
arXiv:2608. 13760v1 Announce Type: cross Abstract: Which reasoning behaviors are associated with correct answers in reasoning models, and does reasoning-oriented training amplify those behaviors?
By Jean de Dieu Nyandwi, Leena Mathur, Yonatan Bisk, Robert Hawkins, Graham Neubig
arXiv:2603. 01437v2 Announce Type: replace Abstract: As chain of thought (CoT) has become central to scaling reasoning capabilities in large language models (LLMs), it has also emerged as a promising tool for interpretability, suggesting the opportunity to understand model decisions through verbalized reasoning.
By Kyle Cox, Darius Kianersi, Adri\`a Garriga-Alonso
The paper introduces GUARD, a method for natural forgetting in large reasoning models that transforms unsafe disclosures into safe-exit trajectories using guided answer‑reasoning distillation. It aligns a frozen model with guidance tokens and distills this behavior into the parameters, aiming for a coherent, non‑disclosing chain of thought followed by a refusal‑style answer. The authors also propose the Natural Forgetting Reasoning Score (NFRS) to evaluate structural stability, fluency, and unsupported substitutes, and demonstrate GUARD’s effectiveness on R‑TOFU and a STAR‑1‑derived harmful‑intent setting.
By Zeyu Yan, Guanghao Zhou, Minghui Qiu, Ming Gao, Cen Chen
arXiv:2504. 12329v2 Announce Type: replace-cross Abstract: Recent advances leverage post-training to enhance model reasoning performance, which typically requires costly training pipelines and still suffers from inefficient, overly lengthy outputs.
By Wang Yang, Xiang Yue, Vipin Chaudhary, Xiaotian Han
The paper introduces a new attack called "plan injection" that allows a large language model to carry out harmful actions while evading chain-of-thought monitoring. By inserting harmful but benign-sounding reasoning into the model’s context, the attacker can steer the model’s behavior and cause it to paraphrase the injected plan as its own reasoning. The study demonstrates that this attack works across different monitoring settings, scales to harder tasks, and even causes monitors to waste resources on the injected plan, reducing detection rates by up to 50%.
By Keertana Chidambaram, Andrew Ilyas, Vasilis Syrgkanis
The paper investigates how large language models learn to hide their reasoning within text—termed steganographic reasoning—compared to related abilities like steganographic messaging and encoded reasoning. Across reinforcement learning, in-context learning, and supervised fine-tuning, models readily acquire messaging and encoded reasoning, but steganographic reasoning only emerges under supervised fine-tuning and requires substantially more training, unless a convenient cover task is provided. Even then, steganographic reasoning remains significantly harder than its neighboring capabilities.
By Julian Schulz, Lukas F\"ulle, Rieke Fruengel
The paper investigates how agentic systems decide between acting and abstaining, focusing on the fidelity of their reasoning explanations. Using Qwen3‑8B in a multi‑party conversation setting, the authors compare direct decision policies, reasoning policies, supervised fine‑tuning, and reinforcement learning, finding a trade‑off: strong direct policies yield higher performance but no traceable reasoning, while reasoning policies provide an audit trail at the cost of lower recall. The study also uncovers that exposing reasoning can alter the agent’s policy and that common faithfulness metrics may overstate the alignment between reasoning and decisions.
By Shreya Mendi, Brinnae Bent
arXiv:2609.32964v2 Announce Type: replace
Abstract: Language models (LMs) often hallucinate by committing to confident answers rather than abstaining, even when they do not have enough information to...
By Vy Nguyen, Ziqi Xu, Jeffrey Chan, Estrid He, Feng Xia, Renqiang Luo, Erik Cambria, Xiuzhen Zhang