Shawn Wang: evaluation

Transcript context

…One thing, yeah, so, so, you know, this came from the Jan recap, which, and diffusion transformers were not really a discussion, and then, obviously, they blow up in February. Yeah. I don't think they're, it's a mixed architecture in the same way that Stripe Tiena is mixed there's just different layers taking different approaches. Also I think another one that I maybe didn't call out here, I think because it happened in February, was hourglass diffusion from stability. But also, you know, another form of mixed architecture. So I guess that is interesting. I don't have much commentary on that, I just think, like, we will try to evolve these things, and maybe one of these architectures will stick and scale, it seems like diffusion transformers is going to be good for anything generative, you know, multi modal. We don't see anything where diffusion is applied to text yet, and that's the wild card for this category. Yeah, I mean, I think I still hold out hope for let's just call it sub quadratic LLMs. I think that a lot of discussion this month actually was also centered around this concept that People always say, oh, like, transformers don't scale because attention is quadratic in the sequence length. Yeah, but, you know, attention actually is a very small part of the actual compute that is being spent, especially in inference. And this is the reason why, you know, when you multiply, when you, when you, when you jump up in terms of the, the model size in GPT 4 from like, you know, 38k to like 32k, you don't also get like a 16 times increase in your, in your performance. And this is also why you don't get like a million times increase in your, in your latency when you throw a million tokens into Gemini. Like people have figured out tricks around it or it's just not that significant as a term, as a part of the overall compute. So there's a lot of challenges to this thing working. It's really interesting how like, how hyped people are about this versus I don't know if it works. You know, it's exactly gonna, gonna work. And then there's also this, this idea of retention over long context. Like, even though you have context utilization, like, the amount of, the amount you can remember is interesting. Because I've had people criticize both Mamba and RWKV because they're kind of, like, RNN ish in the sense that they have, like, a hidden memory and sort of limited hidden memory that they will forget things. So, for all these reasons, Gemini 1. 5, which we still haven't covered, is very interesting because Gemini magically has fixed all these problems with perfect haystack recall and reasonable latency and cost. all these reasons, Gemini 1. 5, which we still haven't covered, is very interesting because Gemini magically has fixed all these problems with perfect haystack recall and reasonable latency and cost. So that's super interesting. So the wildcard I put in here if you want to go to that. I put two actually. One is text diffusion. I think I'm still very influenced by my meeting with a mid journey person who said they were working on text diffusion. I think it would be a very, very different paradigm for, for text generation, reasoning, plan generation if we can get diffusion to work. For text. And then the second one is Dowie Aquila's contextual AI, which is working on retrieval augmented language models, where it kind of puts RAG inside of the language model instead of outside. Yeah, there's a paper called Retro that covers some of this. I think that's an interesting thing. I think the The challenge, well not the challenge, what they need to figure out is like how do you keep the rag piece always up to date constantly, you know, I feel like the models, you put all this work into pre training them, but then at least you have a fixed artifact. These architectures are like constant work needs to be done on them and they can drift even just based on the rag data instead of the model itself. Yeah,…

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