SOURCE-LINKED INTELLIGENCE

Prefilling the Reasoning Channel: Output-Prefix Attacks on Reasoning LLMs

arXiv · AI, language, vision and robotics · article · Sep 24, 2026 · UTC

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 be conditioned on it. This output-prefix attack technique is a cheap black-box prompt injection. Prior work has shown this type of attack can reliably jailbreak non-reasoning models. Most reasoning models add an intermediate scratchpad reasoning step before the assistant's final response. The ability to edit this reasoning channel is exposed by some APIs and attack vectors can be leveraged for rea

Read original source ↗ Open in workspace

recordType

paper

region

Global