The paper investigates a new attack method called output‑prefix attacks on reasoning LLMs, where an attacker prepends a malicious text to the model’s output, thereby conditioning all subsequent tokens on that prefix. The study systematically isolates the scratchpad reasoning channel as a vulnerable vector and compares three attack types—reasoning‑only, output‑prefix‑only, and combined reasoning‑plus‑output‑prefix—across both exposed and hidden reasoning models. Experiments on three 2026‑era frontier models (Gemini 3 Flash Preview, DeepSeek V4 Flash, and Claude Haiku 4.5) show that reasoning alone is largely ineffective, but adding a trivial output prefix can raise attack success rates to as high as 99% for some models, with contextual prefixes outperforming static ones and susceptibility varying by model.
By Luk\'a\v{s} Br\r{u}na, Robert Bridges, Adam Ek
arXiv:2608.30041v1 Announce Type: cross
Abstract: Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later pri...
By Wujie Xiong, Rabimba Karanjai, Yang Lu, Weidong Shi, Lei Xu
The paper introduces a privacy‑preserving zk‑SNARK audit framework that uses adversarial‑style probes to detect logit drift between an approved large language model and a modified deployment. It offers three probe families—token‑based (black‑box), embedding‑based (gray‑box), and stress probes (partial white‑box)—allowing users to balance sensitivity, access, and cost. Experiments across LLM architectures and GPU platforms show token‑based probes achieve the highest mean sensitivity while remaining practical in a black‑box setting, with Groth16 proving times scaling modestly from 1.02 to 1.78 seconds and constant proof size.
By Cameron Wilding, Mina Shaker, Fatemeh Ganji
arXiv:2605. 17062v3 Announce Type: replace-cross Abstract: Spracklen et al.
By Aleksandr Churilov (Independent Researcher)
Large Language Models (LLMs) consume and produce a single sequence of text; hence, if text can be added to the beginning of the LLM's response, i.e., an output prefix, then all subsequent tokens will...
arXiv:2606. 03606v1 Announce Type: cross Abstract: Large language models achieve strong performance on arithmetic reasoning benchmarks, and one common response to arithmetic brittleness is to delegate computation to code.
By Malia Barker, Bishal Lakha, Edoardo Serra, Francesco Gullo
CESBench is a new benchmark for evaluating large language models on cryptographic engineering security for IoT devices, comprising 380 expert‑written items across six sub‑domains such as side‑channel, fault injection, and implementation. The benchmark includes four task types—multiple‑choice, judgment, scenario, and code—each designed to test different competences, with automatic scoring for the first two and LLM‑based judging for the latter two. Evaluation of 11 open‑weight and proprietary LLMs shows strong performance on multiple‑choice and code tasks but weaker results on judgment and scenario tasks, highlighting gaps in justifying security verdicts.
By Wenquan Zhou, An Wang, Jing Liang, Peien Feng, Jingqi Zhang, Yaoling Ding, Liehuang Zhu
The paper introduces the SAST-IR framework to evaluate large language models’ robustness against persuasion attacks in a memory‑less setting, revealing a flaw called "Refusal Inertia" that masks true vulnerability. Using the CP‑Agent and a custom CounterFact‑Strict dataset, the authors demonstrate that simple, diverse attack strategies achieve a 96% success rate, while complex attacks often trigger defensive compliance. The study highlights severe brittleness in current state‑of‑the‑art models when deprived of conversation history.
By Zhuoang Cai
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
Large language models achieve strong performance on arithmetic reasoning benchmarks, and one common response to arithmetic brittleness is to delegate computation to code. Yet models are still often used in settings where they must reason directly from natural language, and trustworthy models should solve small-number arithmetic word problems without external tools.
arXiv:2607. 20216v1 Announce Type: cross Abstract: Malware analysis demands rapid interpretation of complex detonation reports spanning filesystem, network, and process behaviours.
By Adel ElZemity, Shujun Li, Budi Arief
arXiv:2510. 01529v3 Announce Type: replace Abstract: Ball et al.
By Jaiden Fairoze, Sanjam Garg, Keewoo Lee, Mingyuan Wang