Rethinking Pre-Training for Agentic AI with Aakanksha Chowdhery - #759

For the longest time we were measuring, pre training on static benchmarks. If we want these models to be useful as agents, they need to be able to interact with environments. And when we start caring about those agentic tasks, pretraining needs to rethink from fundamentals. This is not just a post training problem to achieve these set of capabilities that we want in the next generation of models. And the kind of benchmarks we need for measuring this kind of intelligence is sometimes not available today. Alright, everyone. Welcome to another episode of the TwiML AI podcast. I'm your host, Sam Charrington. Today, I'm joined by Akansha Choudhury. Akansha is a member of technical staff at Reflection. Before we get going, be sure to hit that subscribe button wherever you're listening to today's show. Kansha, welcome to the podcast. Thank you, Sam. You have a really interesting background. You've trained some of the earliest large language models, including POM and, Gemini one dot o, Gemini one dot five. Tell us a little bit about those experiences. I got into large language models at Google while building one of the distributed systems, that led to the training of, PALM, which is which was our large our largest language model at the time. It had five forty billion parameters. And, people stopped publishing the number of parameters after that. One and, that led me to be, at the forefront of pretraining solving one set of problems after the other that come with scale in first two generations of palm models and then the first two generations of Gemini models. And I think the one thing that you learn when you do pretraining is that at scale, every problem magnifies and things go wrong at every possible part of the stack. So it's always fun and it's always exciting. And you said you were on the infrastructure side? I I was on the ML side, but I've done both. One thing that is super interesting about, pretraining is that you have to be able to think across the stack. So, otherwise, you if you're gonna train something for two or two and a half months, you need to be able to think about every single part of the system. Nice. And so tell us about reflection. What is reflection focused on? So the mission of Reflection is to build frontier open, intelligence, for agentic capabilities. And, the company has been focused on building, post training stack for agentic tasks. And with the most recent fund raise, we are, doing training end to end. So we're building, different tier open agentic models, which are both pretrained and post trained in house. And that's really gonna be a a focus of what we're talking about today, some of the reasons why you think pretraining a different approach to pretraining is key? Is that kinda the way you think about it? Well, the way I really put it is that, for the longest time, we …