, 11:30–12:30
From Parallel Token Prediction to Scalable Language Models
Prof. Stephan Mandt (UC Irvine) on generating several mutually consistent tokens in one forward pass — models pretrained to 3 billion parameters that approach autoregressive quality with inference up to 5.7× faster. Hosted by Prof. Julia Vogt.
- Dates
- Time
- 11:30–12:30
- City
- Zurich
- Venue
- ETH AI Center Foyer, OAT X 11, Andreasstrasse 5, Oerlikon
- Organiser
- ETH AI Center
- Format
- Meetup
- Cost
- Free (registration required)
- Topics
- AI research, LLM efficiency, Open-source LLMs
- Verified via
- ai.ethz.ch