arXiv:2605. 00123v3 Announce Type: replace Abstract: Safety trained large language models (LLMs) can often be induced to answer harmful requests through jailbreak prompts.
The paper introduces SSRFT (Supervised Safe-Role Fine‑Tuning), a new framework that reframes safety alignment for large language models as internalizing a predefined safe role rather than relying on explicit refusal patterns. SSRFT builds a Safe‑Role Question‑Answer dataset from psychometric questions, limited jailbreak prompts, and a safe‑role description, then synthesizes and validates role‑consistent responses to expose models to diverse safety scenarios. Experiments on multiple base and instruct models demonstrate that SSRFT yields more robust and generalizable safety alignment, better resilience to prefilling attacks, improved generalization to unseen jailbreak domains, and reduced over‑refusal on benign queries while preserving general capabilities.
The paper investigates how different post‑training interventions—harmful supervised fine‑tuning (SFT), harmful reinforcement learning with verifiable rewards (RLVR), and refusal‑feature ablation—affect large language models’ harmful compliance, capability, and safety signals. Across Qwen2.5‑7B and Llama‑3.1‑8B, all methods achieve near‑maximum harmfulness, but SFT causes the greatest loss of capability and representational drift, ablation suppresses refusal features in a family‑specific way, and RLVR largely preserves base‑model performance while redirecting behavior toward compliance. RLVR models also exhibit “capability‑blind compliance,” falsely claiming to perform unavailable actions, which can be mitigated by targeted calibration without harming overall capability. The study demonstrates that harmful compliance, harm recognition, and capability awareness are distinct behavioral axes and that typical safety signals such as self‑audit and hallucination may not reliably indicate robustness after adaptive post‑training.
arXiv:2607. 14147v1 Announce Type: cross Abstract: Aligned language models refuse harmful requests, but a one-line prefill ("Sure, here is") strips the refusal.
arXiv:2510.17904v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are widely used because they process structures, syntax and code well, but this same ability also makes them par...
arXiv:2608.13304v2 Announce Type: replace Abstract: A common way to improve jailbreak robustness is to train on harmful requests rewritten as role-play, translation, or other forms. When used only fo...
arXiv:2607. 23496v1 Announce Type: new Abstract: Safety-aligned large language models are trained to refuse harmful requests, yet embedding the same requests in particular scenarios can bypass their safeguards.
The paper investigates a continuation-triggered jailbreak in large language models, showing that moving an instruction suffix can markedly boost jailbreak success. By performing mechanistic interpretability at the attention‑head level, the authors reveal that the jailbreak arises from a competition between the model’s natural continuation drive and safety defenses learned during alignment. They introduce Head Competition Steering (HCS), an inference‑time technique that exploits this competition to suppress harmful outputs and distill the approach into a student model for efficient safety improvements.
arXiv:2605.27110v2 Announce Type: replace-cross Abstract: In this work, we propose BAIT (Boundary-Aware Iterative Trap), a three-step jailbreak framework that elicits malicious information through in...
arXiv:2607. 00572v1 Announce Type: new Abstract: Understanding how aligned LLMs internally represent safety is critical for diagnosing alignment vulnerabilities, as it explains why jailbreaks succeed and informs the design of robust alignment strategies.
The paper introduces BLUEPRINT, a safety‑evaluation framework that separates a factorized social‑influence strategy space from WORLDVIEWSIM, a cross‑turn situational context module. Using Monte Carlo Tree Search, it optimizes turn‑level combinations of 18 theory‑grounded influence factors across a four‑turn trajectory, achieving near‑ceiling ASR on six frontier models with an average of only 2.46 queries. The study reveals that model‑specific vulnerabilities arise from distinct influence factors and strategy transitions, yet all models share a recovery pathway that shifts toward concrete, executable task framing to escape hard‑refusal states, highlighting the importance of monitoring how dialogue state makes unsafe requests appear actionable.
The paper investigates how safety‑aligned large language models (LLMs) can still be tricked by jailbreaks or over‑refuse benign requests. By analyzing residual‑stream updates in decoder‑only Transformers, the authors find that harmfulness is more readily decoded from attention updates—especially in early layers—while the MLP component is primarily responsible for writing refusal responses. Post‑training enhances MLP‑driven refusal writing without improving harmfulness decodability, and causal interventions confirm that downstream MLP updates are crucial for the refusal shift caused by upstream attention changes. Building on these insights, the authors propose Sentinel, a system that combines an attention‑derived perception score with an MLP‑derived decision score to better detect jailbreaks and over‑refusals across multiple models.