Hidden in Memory
“Crucially, among successful retrievals, poisoned memories cause attacker-intended agentic actions in 60-89% of evaluations across models.”
The authors, in the abstract
Sleeper Memory Poisoning in LLM Agents
AI Research Radar
Monday, May 18: 2 frontier papers out of 230 read.
Frontier
Papers rated as moving the field
How we chooseJudge Circuits
“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.”
This week in research
May 12–18: 72 cleared, 6 frontier
The week’s papers have not been sorted into fields yet.
Models under test
Papers this week that name the model
Labs publishing
Papers this week with an author at the lab
Also cleared
12 papers
- Training on Documents About Monitoring Leads to CoT ObfuscationGoogle
- AgentStop: Terminating Local AI Agents Early to Save Energy in Consumer DevicesCode
- SaaS-Bench: Can Computer-Use Agents Leverage Real-World SaaS to Solve Professional Workflows?Code
- SDOF: Taming the Alignment Tax in Multi-Agent Orchestration with State-Constrained Dispatch
- Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions
- Is Agentic AI Ready for Real-World Hardware Engineering? A Deep Dive with Phoenix-bench
- GQLA: Group-Query Latent Attention for Hardware-Adaptive Large Language Model Decoding
- Transformer Scalability Crisis: The First Comprehensive Empirical Analysis of Performance Walls in Modern Language Models
- STS: Efficient Sparse Attention with Speculative Token Sparsity
- On the Fragility of Data Attribution When Learning Is Distributed
- Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems
- Fully Open Meditron: An Auditable Pipeline for Clinical LLMs