Goblin
News
AI news, distilled.
|
News
About
News
About
Filtered by:
skyrl
Clear
Titles
Summaries
7
Mercor and SkyRL Release Full 397B RL Training Recipe for Knowledge Work Agents, Achieving 70% Pass@1 Gain
Research
1
Sep 3
7
Mercor and SkyRL Release Full 397B RL Training Recipe for Knowledge Work Agents, Achieving 70% Pass@1 Gain
Research
· 1 src · Sep 3
Discuss
7
RL Fine-Tuning Enables Small 4B Models to Match Large LLMs as Recursive Agents
Research
1
May 13
7
RL Fine-Tuning Enables Small 4B Models to Match Large LLMs as Recursive Agents
Research
· 1 src · May 13
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
