The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 562
🩲 Super GPU
Blanc Swan
blancsw
AI & ML interests
ChatBot
Recent Activity
View all activity
Organizations
upvoted 2
papers 14 days ago
upvoted a paper 27 days ago
upvoted 7
papers about 1 month ago
Scaling Automatic Research Agents via World Models
Paper • 2608.12564 • Published • 489
NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness
Paper • 2609.08183 • Published • 321
Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills
Paper • 2609.02749 • Published • 408
Language Models Can Control Their Own Attention
Paper • 2609.02737 • Published • 80
Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement
Paper • 2608.31046 • Published • 100
On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability
Paper • 2608.30320 • Published • 62
StudentSim: Training LLM-based Student Simulators
Paper • 2609.01591 • Published • 480
upvoted 3
papers 2 months ago
upvoted a paper 3 months ago
upvoted 2
articles 3 months ago
Article
Kimi K3 Model Overview: 2.8T Parameters, MXFP4 Quantization, and What the Open Weights Mean for the Community
ResterChed
• • 197 Article
Native-speed vLLM transformers modeling backend
hmellor, lysandre
• • 77 upvoted a paper 3 months ago
upvoted a paper 4 months ago
upvoted a collection 6 months ago
upvoted a paper 6 months ago