Joseph

lkdsmr's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted a paper about 12 hours ago

Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training

Paper • 2605.12483 • Published 1 day ago • 7

updated a collection 28 days ago

papers

Collection

4 items • Updated 28 days ago

upvoted a paper 28 days ago

TIP: Token Importance in On-Policy Distillation

Paper • 2604.14084 • Published 29 days ago • 15

updated a collection about 2 months ago

papers

Collection

4 items • Updated 28 days ago

upvoted a paper about 2 months ago

IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse

Paper • 2603.12201 • Published Mar 12 • 53

upvoted 3 papers 2 months ago

On-Policy Self-Distillation for Reasoning Compression

Paper • 2603.05433 • Published Mar 5 • 9

Not all tokens are needed(NAT): token efficient reinforcement learning

Paper • 2603.06619 • Published Feb 20 • 1

Overconfident Errors Need Stronger Correction: Asymmetric Confidence Penalties for Reinforcement Learning

Paper • 2602.21420 • Published Feb 24 • 6

updated a collection 2 months ago

papers

Collection

4 items • Updated 28 days ago

upvoted 2 papers 2 months ago

PACED: Distillation at the Frontier of Student Competence

Paper • 2603.11178 • Published Mar 11 • 4

Spatial-TTT: Streaming Visual-based Spatial Intelligence with Test-Time Training

Paper • 2603.12255 • Published Mar 12 • 91