Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Models
Paper • 2608.16647 • Published • 15
AI & ML interests
None defined yet.
Papers
Diff4Splat: Controllable 4D Scene Generation with Latent Dynamic Reconstruction Models
Detecting Data Contamination from Reinforcement Learning Post-training for Large Language Models
DeyangKong
authored a
paper 20 days ago
DeyangKong
submitted a
paper to Daily Papers 22 days ago
ZiangWu
authored a
paper about 1 month ago
ZiangWu
submitted a
paper to Daily Papers about 1 month ago
skysss
authored 3
papers about 1 month ago
Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Model
Paper • 2607.22083 • Published • 11
Beyond Superficial Forgetting: Thorough Unlearning through Knowledge Density Estimation and Block Re-insertion
Paper • 2511.11667 • Published
PACE: Prefix-Protected and Difficulty-Aware Compression for Efficient Reasoning
Paper • 2602.11639 • Published
wyzmike
authored a
paper 6 months ago
skysss
authored 2
papers 7 months ago
DeyangKong
authored a
paper 7 months ago
skysss
authored a
paper 7 months ago
DeyangKong
authored 2
papers 7 months ago
skysss
authored 3
papers 7 months ago
CoSineVerifier: Tool-Augmented Answer Verification for Computation-Oriented Scientific Questions
Paper • 2512.01224 • Published
Nanbeige4-3B Technical Report: Exploring the Frontier of Small Language Models
Paper • 2512.06266 • Published • 8
From Failure to Mastery: Generating Hard Samples for Tool-use Agents
Paper • 2601.01498 • Published • 2
DeyangKong
authored a
paper 7 months ago
DeyangKong
submitted a
paper to Daily Papers 7 months ago
wyzmike
authored a
paper 8 months ago