Wednesday, November 4: 3 frontier papers out of 433 read.
Papers rated as moving the field
How we choose“We measure this at the generation stage across six models and ten consumer-service domains, for three attack families, and show that it transfers to production browser agents, outperforming instruction injection, and to a retrieval pipeline (1.0-57.0% end-to-end success, depending on the retriever).”
The authors, in the abstract
Weaponized Safety Warning Mechanisms for Controlling AI Agents
BrainwashedLLM achieves ASR@5 between 0.74 and 0.98 across four state-of-the-art LRMs without querying targets during attack construction.
PrimeSim leads four of five WorldSimProbe suites and posts the best overall EWMBench score against five open-source action-conditioned world models.
Oct 29 – Nov 4: 288 cleared, 20 frontier
Papers this week that name the model
Papers this week with an author at the lab
48 papers, by field