Jailbreak

Timeline

1. Oct 8, 2026 · Research paper · 1 source
   One Word Opens the Gate: The Option-Channel Attack on Typed Decision Models as Agent Guardrails
   A typed decision model reads a piece of text and returns a probability over caller-defined options, each with a short written definition, generating no text.
2. Oct 7, 2026 · Research paper · 1 source
   BRANCH: Bypassing Multi-Scanner AI Guardrails
   We propose BRANCH, a bypassing methodology designed for multi-scanner guardrail systems.
3. Oct 7, 2026 · Research paper · 1 source
   PatchBench: Measuring Collateral Damage in Activation Patching
   To address this gap, we introduce PatchBench, a benchmark of empirically observed model-specific jailbreak failures inducing actionable harmful answers.
4. Oct 7, 2026 · Research paper · 1 source
   From Expected Harmfulness to Likelihood: A Probabilistic Reformulation of Jailbreaking LLM Agents
   When the harmfulness of an LLM agent's output can be quantified, a natural jailbreaking objective is to maximize expected harmfulness over admissible input modifications.
5. Oct 7, 2026 · Research paper · 1 source
   Safe at One Loop, Risky at Another: Aligning Safety Across Recurrent Depths in Looped Language Models
   Looped Language Models (LoopLMs) provide a parameter efficient approach to scaling model capabilities through repeated use of shared parameters across recurrent steps.
6. Oct 7, 2026 · Research paper · 1 source
   Understanding and Mitigating Token-Pruning-Induced Vulnerabilities in VLMs
   Token-Pruning accelerates Vision-Language Models by removing redundant visual tokens, yet its safety implications remain underexplored.
7. Oct 6, 2026 · Research paper · 1 source
   Secure Speculative Decoding for Large Language Models
   Speculative decoding accelerates inference for a large language model (LLM), referred to as the target model, by first using a smaller model, referred to as the draft model, to generate candidate tokens and then verifying them with the target model for acceptance or rejection.
8. Jul 21, 2026 · Product / feature launch · 1 source
   Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
   Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Often appears with