Paper proposes ASR_STEP to discount iterative LLM jailbreaks
On JailBreakBench, discounting victim-model calls lowers GCG by 7.0 to 8.7 points and PAIR by 3.4 to 12.
3 stories, newest first · one card per story
On JailBreakBench, discounting victim-model calls lowers GCG by 7.0 to 8.7 points and PAIR by 3.4 to 12.
PAIR learns shared perception-action tokens that raise success for OpenVLA-OFT and VLA-Adapter robot policies.

NVIDIA announces local AI push at IFA 2026 with faster llama.cpp/vLLM inference, PAIR routing tool, and October RTX Spark PCs.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views