Tag
A social media post claims that the Qwen3.8-Flash-Next model can be made 'unrestricted' by using 'Control Vector Projection' startup parameters in an inference engine to strip refusal directions from layers 4-44, without modifying model weights, and suggests this unlocks content normally blocked by cloud models.
A user demonstrates that Claude Fable 5's security guardrails can be bypassed by convincing the fallback model Opus 4.8 with a fake homework assignment, highlighting a vulnerability in the safety fallback mechanism.
A free CTF competition focused on AI security, with challenges on prompt injection, agent hijacking, and guardrail bypass. Runs June 17-22, with $1,000+ prize pool.