On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training
Paper • 2609.36659 • Published • 87
Mila Lu
milalu7385
AI & ML interests
None yet
Recent Activity
upvoted a paper 5 days ago
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training upvoted a paper 11 days ago
Raven: The Harness of Harnesses for Composable Agentic Intelligence upvoted a paper 11 days ago
MassAlloc Attention: Let Attention Allocate Its Own ComputeOrganizations
None yet
upvoted a paper 5 days ago
upvoted 3
papers 11 days ago
Raven: The Harness of Harnesses for Composable Agentic Intelligence
Paper • 2609.33439 • Published • 674
MassAlloc Attention: Let Attention Allocate Its Own Compute
Paper • 2609.32712 • Published • 75
CoWindow Attention: Full Causal Coverage Is a Collective Property
Paper • 2609.32704 • Published • 67