I Built Non-Autoregressive Decision Models a Year Ago. Then a Frontier Lab Called It a "Breakthrough"
40 points by Yogthos
40 points by Yogthos
Should be merged with https://lobste.rs/s/ojukrw/laya_33ms_multilingual_system_1_decision, I think
I haven’t been in the trade for 20 years but, in my days, I solved the "decision problem" with simple bayesian learning. When the input is text to classify and the number of classification is low, bayesian are really really good.
I can be wrong, but my impression is that they are building a generic classifier, for which you supply the task in a natural language and the schema of the output, and get some (hopefully, good) predictions.
Classifiers are obviously not a new thing, but the old way is to gather a labeled dataset specifically for your task and then train a model for it. For Jev, if I understand it correctly, you don't need to train it for each new task.
In my March 2025 work, I used frozen sequence embeddings combined with a separate PPO value network. It worked for turn-by-turn sales prediction, but it was not end-to-end and could not handle dynamic new questions at runtime.
(emphasis mine)
Maybe this is just my lack of ML knowledge showing, but isn’t this kind of the entire selling point with Jev? I was intrigued but when I got to this part of the article I couldn’t help but feel that the title is a bit misleading… wouldn’t your model be closer to a traditional classifier then? I’m sure there’s more to it than that, but I need someone to ELI5.