Cohere Labs talk: changing model architecture late in training instead of freezing after pretraining
Cohere_Labs · x · 2026-11-03
Cohere Labs released a talk by research engineer Anupam Nayak on transforming AI model architectures efficiently: rather than keeping the architecture fixed after pretraining, it can be modified late in training. Full talk at the link. The direction has practical implications for cutting training costs and reusing existing training progress.
More from Research
- Graph Isomorphism Algorithm Improved to n^O(log n), Exponent Drops from 2 to 1 — fortnow · 2026-11-03
- Researchers Wrote a Paper Assessing Matt Pocock's /retro Skill for Coding Agents — mattpocockuk · 2026-11-03
- Anthropic Releases FlashProve Prototype for Provable LLM Inference — scaling01 · 2026-11-03
- Why AI cracking Navier-Stokes doesn't mean it can auto-discover better models — snikolov · 2026-11-03
- STEP: surgical point tracking from sparse endpoints cuts error from 31.5 to 16.8 pixels — ftm_guney · 2026-11-03
- Allen Institute's OpenScope transforms into Global Brain Observatory, sharing brain-wide recordings openly — KordingLab · 2026-11-03