Thinking-Free Policy Initialization Makes Distilled Reasoning Models More Effective and Efficient Reasoners
Paper
•
2509.26226
•
Published
•
32
Kai Yang
yangkaiSIGS
AI & ML interests
None yet
Recent Activity
updated
a Space
4 days ago
yangkaiSIGS/entropic
authored
a paper
21 days ago
Thinking-Free Policy Initialization Makes Distilled Reasoning Models
More Effective and Efficient Reasoners
authored
a paper
21 days ago
EntroPIC: Towards Stable Long-Term Training of LLMs via Entropy Stabilization with Proportional-Integral Control
Organizations
authored
9
papers
21 days ago
EntroPIC: Towards Stable Long-Term Training of LLMs via Entropy Stabilization with Proportional-Integral Control
Paper
•
2511.15248
•
Published
•
6
Exploration and Anti-Exploration with Distributional Random Network Distillation
Paper
•
2401.09750
•
Published
A Two-stage Reinforcement Learning-based Approach for Multi-entity Task Allocation
Paper
•
2407.00496
•
Published
BATON: Aligning Text-to-Audio Model with Human Preference Feedback
Paper
•
2402.00744
•
Published
Novelty-Guided Data Reuse for Efficient and Diversified Multi-Agent Reinforcement Learning
Paper
•
2412.15517
•
Published
Exploration by Random Distribution Distillation
Paper
•
2505.11044
•
Published
Novelty-based Sample Reuse for Continuous Robotics Control
Paper
•
2410.13490
•
Published
CDSA: Conservative Denoising Score-based Algorithm for Offline Reinforcement Learning
Paper
•
2406.07541
•
Published
authored
a
paper
about 2 years ago