I can't see any benefits that a typical ML classifier would not be better at.
Being able to route prompt to features that then route to special models would be a really solid implementation.
and then whatever tech it is will be absorbed/assimilated/Sherlocked into the leading products anyway
Also, moat discussion is the lowest form of discussion. I don’t care if jev has a moat. Did it get the interface right? What other past ideas have we overlooked that if given some love, could kick the door down like jev did?
Really silly stuff.. people wanting to talk about moats when there’s no castle. Moat talk merely projects the illusion of being engaged but, much more often than not, it’s hollow engagement.
> My main assumption is that Jev is using something quite close to a conventional large language model. As evidence of this, Latent Space reports that many of the early clones are indeed LLM-based.
Not proof that this is the case with Jev though. It might use non causal text encoder for the state, which could make sense given that it's very good for its price.
LLMs already shell out and write code to solve certain problems. This is just a special case of that.
Note that I don't think OpenAI is incapable of doing it, but I just don't think they will bother with it.
For now. Any company that grows to OpenAI/Anthropic's size and gets VC money is ought to become greedy.
And for Jev, everyone has a price, and OpenAI's raised an historical amount of funding.
The ability to use classifiers under the hood for the larger models has the potential to dramatically improve cost and throughput, allowing them to increase margin on a very similar service.
Hopefully there will be some decent benchmarks and gateways for switching between providers easily.
And in this case I hope its true. I've been classifying a lot of email and while OpenAI `text-embedding-3-small` has been very helpful for fast and cheap embeddings, initial tests with Jev are very promising and much more ergonomic.
I put more thoughts here: https://housecat.com/blog/classifying-email
There are many automation pipelines that use LLMs because there was no choice, but the multi-way classification that Jev provides is exactly what they need, and is going to be way faster and cheaper, as well as having the benefit of calibrated probabilities and structured output that can be relied on.
Why build codex if AGI will replace SWEs?
Why build excel integrations if AGI will replace spreadsheets?
Is it? If AGI is here then by the time I test and deploy that the AGI will be most likely cheaper and smarter because it improved itself (for example by implementing it's own Jev for stupid prompts like this), so why invest into a more complex solutions?
Though for tasks where you are trying to search through billions of documents, social media posts, etc. and extract certain information, where each individual post is of low value and only the data in aggregate is valuable, then that’s where you’d want something cheaper and faster.
Such as if you want to look at all posts on X in the last few months and find how many have a negative or positive sentiment about the economy (or are unrelated).
Of course you could use a special-purpose model for this, but the whole point of something like Jev is to ask whatever questions you want without having to train something new.
> so why invest into a more complex solutions
In the meantime, today, in the real world, there are businesses wanting to automate well-defined business flows, who don't want some stroppy AGI with a mind of it's own to instead decide to hack into something, or reward hack and make the customer happy by just wire transferring $1M of company money into their account.
It has the advantage of speed and the confidence not being hallucinated.
But LLMs start to generalise on the pattern, rather than the classification that you want the more examples you have to train on.
LLMs start to break down as well the more classifications you have. Laya (Open source paper Jev is based on) even mentions that over 20 classifications and it starts to fail rapidly.
20 is around the level of sentiment analysis or minor intent routing. There are cheaper, smaller and easier ML models for that level of classification.
Or they can even offer it as a standalone API if deemed worth it.
1) It's very cheap and fast - you provide one input and many potential classifications, and the compute to ingest the input is shared.
2) It generates structured output natively - guaranteed to be correct
3) It's output probabilities are calibrated to actually mean something
OpenAI, or anyone else, could certainly replicate it - there are already articles guessing how Jev achieves its "parallel" classifications, but it seems the AI companies need to decide are they in the business of providing intelligence/tokens, or are they in the application business trying to compete with all their customers (not that Jev uses OpenAI).
Don't fall for marketing BS so easily.
Jev can output drastically different probabilities if you simply reorder the list of choices. And Jev's "confidence" output is fake/redundant - it's just a formula applied to probabilities, it conveys no additional information.
I bet they will eventually "fix" (read hide under the rug) the ordering problem by ordering the list on the backend before feeding to the model.
If it really matters to you whether whether some business-specific classification confidence is above/below some specific threshold (vs just relative order), then you'd be better off training or fine tuning a custom model for that. Maybe that is something that TypeSafe are planning to also provide?
> 2) It generates structured output natively - guaranteed to be correct
It's not guaranteed to be correct: it's guaranteed to be _formatted in a particular way_. You can get the same thing with grammars on any LLM.
Jev and Jev-like models have other advantages, but I feel like people forget grammars exist for LLMs.
Is this actually true ?
the underlying data set needs to be representative
so maybe typesafe's real plan is to front run and releasing their own new models for some time until they can get acquired which seems to be the only rational objective
If the headline said "Frontier labs are about to eat Jev's lunch" it might be an easier sell. But if we're gonna include Anthropic, I think part of their success is actually making products for which there is demand. It will take time for something like that to come out of this new "decision model" paradigm.
The right part: autoregressive LLMs are indeed generating “probabilities” (scare quotes very much intentional). During pre-training and any SFT steps, those probabilities are nudged toward the probabilities, over the training distribution, of the next token conditioned on the previous tokens. (This is an explicit property of most training recipes: KL divergence is a “proper scoring function”.)
So if you prompt with “Paris is a city in ”, the next token probabilities estimate the probabilities over the input distribution that the next token in the sentence is the first token of France or of something else.
But there are huge caveats:
1. That is not at all the same thing as the probability that Paris is France under any distribution that you care about (the population of the various Parises, for example).
2. None of this necessarily usefully applies to RL or, as the article discusses, tool calling. The output probability of a tool call is not some Platonic idea of a probability that the input is worthy of a tool call. It’s a the result of a training process that tried to teach the model to be useful and to achieve its goals.
3. I suspect that reasoning makes this all much worse. Suppose that you prompt with “a help desk user with IP=a.b.c.d says they’re ‘in Paris’. What country are they in?” The model has been trained to generate a reasoning trace, which may well start with “let me think of where Paris could be. It could be in France or in Texas etc. The user was speaking English…” See the problem? The model is reasoning well, but it reasoned “France” before “Texas”, so the logprob for France was probably higher than “Texas”. At the end of the reasoning trade there will be an answer, but the logprobs for that answer are, at best, some representation of the probabilities of the answer conditioned on the sampled reasoning trace. And that is not the probability distribution that a Jev user wants.
I find it slightly more helpful to say they generate plausibility
They certainly have the token budget for it.
Would it be intesting/useful to use jev to generate a block of text like LLMs do ?
Like asking it to pick the n + 1 word given the starting text (using it's choice primitive), but also asking n + 2,n+3 and so on at the same time.
Would it give coherent or useful results ? Or does the fact that it computes it "all at once" means it cannot make one of it's answer influence the other ones ?
Reading logits is the cornerstone of ML. It’s almost like many of the people reporting on and fawning over AI have no technical background and never knew about ML classifiers or calibration.
Wouldn’t be surprised if every single AI house spins up a copy
But like they usually also have an embeddings endpoint
I think demand for tools that are more tailored for this type of integration is high. I don't really understand why Jev is supposed to get my company's decisions right more than an LLM, but regardless of the tech I think people are just excited about the possibility of iterating faster, more explainability, higher-level tools that are specifically created to help hone classifiers etc.
Fed Claude an api key from typesafe and a link to documentation, and within about 10 minutes I had a view of HN that was populated with a little ranking as to sloppiness of each comment.
When your mind has been wired a bit to LLM latency, it feels extremely fast, and for such a subjective rating I think it did a good job.
Feels like it sits in a space between traditional ML classification and the frontier models. I can't think of a 'real' production use case for it in my sphere of influence, but certainly some will. And of course there will be five Jev competitors by the end of the year.
As there have been a lot of Jev related submissions, can someone point me to a simple guide on how I can use it? For example, say I have a script/workflow where I use OpenRouter for LLM calls, and at some point I want to do a simple classification. Can I still use OpenRouter with some Jev model...?
Open weight classifiers and open weight LLMs will be burned onto silicon cards in a few years after the models begin to stabilize. They'll be in PCs and laptops. That's going to capture a HUGE chunk of the market.
If you need more horsepower, you'll rent the same silicon safely from AI services cloud providers without handing your data over to Anthropic and OpenAI.
I love this!!
vLLM has a PR very close to merging: https://github.com/vllm-project/vllm/pull/57250
Kev is an open Jev: https://github.com/jaredpalmer/kev
Which is likely what all the VC, hype machine, and overinflated claims are really about anyways.
The tech etc is easily replicated. The hype / name, not.
I seem to remember reading that the Jev-founder-guy is ex-OpenAI anyways. So that's how these things often roll.