Sign up to save your podcasts
Or
Share Abstracts: NeurIPS 2024 with Weizhu Chen
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Abstracts: NeurIPS 2024 with Weizhu Chen
Next-token prediction trains a language model on all tokens in a sequence. VP Weizhu Chen discusses his team’s 2024 NeurIPS paper on how distinguishing between useful and “noisy” tokens in pretraining can improve token efficiency and model performance.
Read the paper
Get the code
View all episodes
By Researchers across the Microsoft research community
4.8
8080 ratings
Next-token prediction trains a language model on all tokens in a sequence. VP Weizhu Chen discusses his team’s 2024 NeurIPS paper on how distinguishing between useful and “noisy” tokens in pretraining can improve token efficiency and model performance.
Read the paper
Get the code
More shows like Microsoft Research Podcast
View all
NVIDIA AI Podcast
341 Listeners
AI Today Podcast
154 Listeners
Practical AI
213 Listeners
Last Week in AI
306 Listeners
Machine Learning Street Talk (MLST)
90 Listeners
Dwarkesh Podcast
506 Listeners
Big Technology Podcast
478 Listeners
NEJM AI Grand Rounds
59 Listeners
No Priors: Artificial Intelligence | Technology | Startups
131 Listeners
Latent Space: The AI Engineer Podcast
95 Listeners
Possible
123 Listeners
The AI Daily Brief: Artificial Intelligence News and Analysis
591 Listeners
Practical: AI & Business News
26 Listeners
AI + a16z
35 Listeners
AI Applied: Covering AI News, Interviews and Tools - ChatGPT, Midjourney, Gemini, OpenAI, Anthropic
136 Listeners