DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data
Paper • 2608.13517 • Published • 32
AI & ML interests
None defined yet.
Papers
DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data
The Arbiter Agent: Continually Monitoring Multi-Agent Conversations to Detect Emergent Misalignment
giannor
submitted a
paper to Daily Papers 3 months ago
filo362
submitted a
paper to Daily Papers 5 months ago
giannor
submitted 3
papers to Daily Papers 5 months ago
PsychoSafe: Eliciting Psychologically-Informed Refusals in Large Language Models
Paper • 2606.09697 • Published • 7
BrainSurgery: Reproducible and Reliable Declarative Weight Manipulations for Model Editing and Upcycling
Paper • 2606.09707 • Published • 8
LLMs Can Leak Training Data But Do They Want To? A Propensity-Aware Evaluation of Memorization in LLMs
Paper • 2606.06286 • Published • 7