The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 259
Kabir Bhat
KABIR-BHAT
·
AI & ML interests
Quantization is my passion
Recent Activity
updated a model about 9 hours ago
KABIR-BHAT/tiny-transformer-multitask published a model about 9 hours ago
KABIR-BHAT/tiny-transformer-multitaskOrganizations
None yet
upvoted 4
papers about 9 hours ago
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents
Paper • 2609.39102 • Published • 141
MaLiang-Harness: A Programmable Path to Image and Video Generation
Paper • 2609.34309 • Published • 384
AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks
Paper • 2609.38288 • Published • 93
upvoted a paper 21 days ago