May 17–23: 3 frontier papers out of 1,620 read.
Papers rated as moving the field
How we choose“A HACR audit on 500 stratified task keys shows that unsupported action-critical claims reach unsafe execution for naive agents (100.0%) and prompt-only defenses (49.6%), but not for ECA.”
The authors, in the abstract
Evidence-Carrying Multimodal Agents
“Crucially, among successful retrievals, poisoned memories cause attacker-intended agentic actions in 60-89% of evaluations across models.”
“We find that judgments across structured understanding and open-ended preference tasks share a sparse, generalized Latent Evaluator sub-graph in the mid-to-late multi-layer perceptrons (MLPs); zero-ablating it collapses judgment while preserving world knowledge in architecturally modular models.”
May 17–23: 50 cleared, 3 frontier
The week’s papers have not been sorted into fields yet.
Papers this week that name the model
Papers this week with an author at the lab
47 papers