Ornith-1.5: From Self-Scaffolding to Self-Improvement
The developers of Ornith-1.5 have released a new suite of foundation models capable of end-to-end self-improvement through task generation and reinforcement learning. The models range from 9B to 397B parameters and demonstrate competitive performance against industry leaders like Claude Opus.
Why it matters
The shift toward self-improving AI models represents a potential leap in autonomous reasoning and coding capabilities, challenging the current dominance of proprietary closed-source models.
Today, we are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement. Ornith-1.5 extends the self-scaffolding framework introduced in Ornith-1.0 into a more complete self-improvement loop: the model proposes new tasks, generates task-specific scaffolds, and produces solution rollouts for reinforcement learning, continuously creating new learning experiences from which it can improve.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in