Aravind Srinivas: recommendation

Transcript context

…Man, there’s a lot of questions there. Is it possible to form that SLM? You can use an LLM to help with the filtering which pieces of data are likely to be useful for reasoning? Absolutely. And these are the kind of architectures we should explore more, where small models… And this is also why I believe open source is important, because at least it gives you a good base model to start with and try different experiments in the post-training phase to see if you can just specifically shape these models for being good reasoners. So you recently posted a paper, A Star Bootstrapping Reasoning With Reasoning. So can you explain chain of thought, and that whole direction of work, how useful is that.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.