-
Training-Free Dynamic Upcycling of Expert Language Models
Paper • 2603.29765 • Published • 10 -
Sharing is Caring: Efficient LM Post-Training with Collective RL Experience Sharing
Paper • 2509.08721 • Published • 665 -
All is Not Lost: LLM Recovery without Checkpoints
Paper • 2506.15461 • Published • 40 -
NoLoCo: No-all-reduce Low Communication Training Method for Large Models
Paper • 2506.10911 • Published • 9
AI & ML interests
We network together the core resource for machine intelligence to flourish alongside human intelligence https://www.gensyn.ai/research
Recent Activity
View all activity
Papers
The Role of Feedback Alignment in Self-Distillation
IR3DE: A Linear Router for Large Language Models
-
Training-Free Dynamic Upcycling of Expert Language Models
Paper • 2603.29765 • Published • 10 -
Sharing is Caring: Efficient LM Post-Training with Collective RL Experience Sharing
Paper • 2509.08721 • Published • 665 -
All is Not Lost: LLM Recovery without Checkpoints
Paper • 2506.15461 • Published • 40 -
NoLoCo: No-all-reduce Low Communication Training Method for Large Models
Paper • 2506.10911 • Published • 9
RL Swarm is an open source system for peer-to-peer gossip-based reinforcement learning over the internet.
models 6
Gensyn/testing
112k • Updated • 4
Gensyn/Qwen2.5-7B-Instruct
Text Generation • 8B • Updated • 41 • • 7
Gensyn/Qwen2.5-72B-Instruct-bnb-4bit
Text Generation • 75B • Updated • 2
Gensyn/Qwen2.5-32B-Instruct-bnb-4bit
Text Generation • 34B • Updated • 6
Gensyn/Qwen2.5-1.5B-Instruct
Text Generation • 2B • Updated • 227 • • 17
Gensyn/Qwen2.5-0.5B-Instruct
Text Generation • 0.5B • Updated • 321 • • 33
datasets 0
None public yet