Transformer
also referred to as: transformers
9 statements across 7 episodes · 4 bullish · 2 bearish · 7 people on the record · first statement Sep 1, 2023 by Guido Appenzeller · said 2 times in 2 episodes since 2019 · across every show →
Mentions by year
brought up most by Marc Andreessen (2)
tap a year for its mentions
2023 1 mention in 1 episode
Everything said about Transformer, oldest first
Sep 1, 2023
Mar 27, 2024 negative
Mar 15, 2025 bullish
Mar 15, 2025
Kumar: Theoretical gains won't unseat transformers without matching years of optimization
“But because of all the engineering work that the community has done around Transformers, it's like, you know, it's very good. And you're not going to just sort of unseat that, you know, just by an idea, right? There's a lot of work to be done.”
Apr 14, 2025 bullish
Yudis: Transformer architectures bring LLM-style data scaling laws to autonomy
“There's a similar you know, let's say paradigm shift happening in autonomy and in vision systems as well with transformers, and so what that ultimately means in autonomy is the similar dynamics in large language models, you can put lots of data in, and you get…”
Jul 16, 2025 bearish
Dec 8, 2025 neutral
Mar 17, 2026 positive
Transformers compute precise Bayesian posteriors down to 10^-3 bits accuracy
“We trained these models and we found that the transformer got the precise Bayesian posterior down to 10 to the power minus three bits accuracy. It was matching the distribution perfectly. So it is actually doing Bayesian in the mathematical sense, given a task…”