Yu li

Yukkkop's activity

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

upvoted a paper about 1 hour ago

Beyond Euclidean Clipping: Overcoming Exploration Collapse in LLM RL via Riemannian Isometric Policy Optimization

Paper • 2607.10169 • Published 19 days ago • 14

upvoted an article about 1 hour ago

view article

Article

LFM2.5-Encoders for Fast Long-Context Inference on CPU

• 

LiquidAI

•

about 22 hours ago

• 45

upvoted a paper 6 days ago

KAT-Coder-V2.5 Technical Report

Paper • 2607.05471 • Published 23 days ago • 10

upvoted 5 papers 9 days ago

What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness

Paper • 2607.08046 • Published 21 days ago • 14

Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Models

Paper • 2605.26895 • Published May 26 • 23

DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning

Paper • 2605.25604 • Published May 25 • 139

Learning to Foresee: Unveiling the Unlocking Efficiency of On-Policy Distillation

Paper • 2605.11739 • Published May 13 • 61

Nudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVR

Paper • 2605.15726 • Published May 15 • 36

upvoted 9 papers 15 days ago

Morphing into Hybrid Attention Models

Paper • 2606.30562 • Published about 1 month ago • 52

Efficient Reasoning via Thought-Training and Thought-Free Inference

Paper • 2511.03408 • Published Nov 5, 2025 • 1

Why Fine-Tuning Encourages Hallucinations and How to Fix It

Paper • 2604.15574 • Published Apr 16 • 26

ESPO: Early-Stopping Proximal Policy Optimization

Paper • 2605.29860 • Published May 28 • 21

Unsupervised Process Reward Models

Paper • 2605.10158 • Published May 11 • 28

Trust Region Policy Distillation

Paper • 2607.04751 • Published 24 days ago • 34

Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE

Paper • 2607.07740 • Published 22 days ago • 24

UP: Unbounded Positive Asymmetric Optimization for Breaking the Exploration-Stability Dilemma

Paper • 2607.06987 • Published 22 days ago • 9

Super Apriel: One Checkpoint, Many Speeds

Paper • 2604.19877 • Published Apr 21 • 3

upvoted a paper 29 days ago

DanceOPD: On-Policy Generative Field Distillation

Paper • 2606.27377 • Published Jun 25 • 81

upvoted 2 papers about 1 month ago

DualPath: Breaking the Storage Bandwidth Bottleneck in Agentic LLM Inference

Paper • 2602.21548 • Published Feb 25 • 56

Tapered Language Models

Paper • 2606.23670 • Published Jun 22 • 10