Not sure on that, maybe the options to choose from will be generated and curated. Same as we do with tagging datasets for images. Might be wildly successful for real world decisions.

Which few to none seem to have understood why, and they do not incorporate, composing their requests with implied information any AI must guess what the hell this request is talking about. Look for and replace implied information with explicit information (that does not have to be detailed, just the correct non-casual language loaded with implied context.)

This is true Jevons Paradox (hence the Jev name) there will be so many usecases, applications and even new jobs out of this.

Learned also that Jev was trained on 100%(!) synthetic data.

What a great time to be alive.

Anthropic's Claude fingerprinting technology at work; randomly inject "real" everywhere. If it was Codex you would have seen load-bearing choice.

Claude insists on injecting the word real or actual everywhere.

I kinda wonder if being trained on other English dialects, particularly Indian English, causes this

It's a specialized classifier model. It classifies input text into categories with a confidence score. Usually those classifiers are small like in the OP but jev is supposedly big, smart, and fast enough to play DOOM by having the scene described in text and classifying it into button presses.

I get the point, and it's nice, but I think the "Jev" naming is confusing (and it could be legally dangerous).

What’s novel is how fast and cheap Jev is while maintaining quality. If they’re trying to say they made the same thing, that is likely incorrect. Getting the same result 100x faster is in fact a breakthrough technology.

Performance for this kind of thing should be best on any hardware that has high prefill speeds. As basically this is "do prefill only, measure scores, skip decode entirely".

I don't know how the Mac stuff compares on that front.

I have the same thing replicated in my own bespoke inference engine (for DGX Spark, in Rust & CUDA) and get answers pretty much as fast as the Jev openrouter endpoint.

https://github.com/rdaum/eider/

It's running over Qwen3.6. Getting it working with Qwen3.8 Flash Next now and getting a battery of tests and examples before I go more public with it.

I think so, yes.

However, it might have fewer restrictions than a BERT and/or is smarter (whatever that means).

Because it's forced to rate them, there's should be a separate uncertainty parameter for both.

Did you try Tabs and Spaces?

You mostly go for „Bar” only after you already went „Foo”.

I really hate the way that LLMS design websites.

Unfortunately huggingface.co is blocked by my company's firewall and VPN so it breaks when downloading a model.

Are there any huggingface mirrors out there?

You sure the answer should have been 1? As a human I'd say I don't have enough information to answer this confidently, but "argument effectively" strongly suggests soft skills to me

you're not alone

As a designer; only slightly. I'm not there to be blown away by awesome design.

For me it's a bit like with some of the LLM prose - uncanny valley territory.

LLM copyedits such as these aren't my idea of clear.

The sites Claude generates by default are almost always in dark mode (no option to switch) and are difficult to read when it comes to font, font color and size choices. It's almost telling you the "author" has zero interest in user experience and doesn't care. This site is several levels above that.

https://ssi.inc/ comes to mind

It's overly automated and repetitive in its styling. Humans make odd stray adjustments to styling manually. LLMs build pages very efficiently. Unless you're very anal-retentive when building a site, there's going to be some distinct flair that isn't just a repeating segment.

It's like it was made by the world's most anal-retentive Wordpress theme builder. They went over it a thousand times until it was perfectly optimized, no distinguishing marks, no stray tiny misalignments, no single-use stylings.

That’s a scary thought (at least, to me). But all change is scary.

The thought is a new wave of people who only know LLM-generated sites, so those design patterns are what they demand/emulate/etc. across the spectrum of user interfaces.

The only previous trend I can draw a parallel to was when Comic Sans and Microsoft Clip Art dominated every flyer and poster.

It's not just you.

This site actually reminds of the TUIs that one uses to install an OS from the text-console. It's not so bad. The prose itself is irritating. The site itself also has some bugs (text overlapping with UI borders for no reason). The lime-green color is a little awkward to my eye, but maybe that's just me (I say this as someone who usually likes lime-green -- maybe the problem is that this site needs _more_ lime-green).

You'd think they'd have addressed verbosity in the last couple of quarters since it's burning their inference at rates they seem to care about. But instead they're more focused on their cyber security FUD distribution so they can lock out all competitive angles possible. I can't wait for the American Greed miniseries.

I tell the agent to outline the greebling, which is the when you add extra details that aren't really necessary.

Initially it feels like the result will be too empty, but once the greebling is removed it most often looks better

I don’t mind the generic dark themed LLM ones even if they all look the same but this particular block looking one is not my fav

Cannot speak for all website, but this one is bad. I was clicking on some text thinking they were tabs or buttons, not the best UX

Isn't this one mimicking the typesafe ai horror website ?

Its why sites that are "well designed" but obviously just use a squarespace or wix template feel so cheap.

Same. I honestly am very satisfied with the aesthetics of free Wordpress blogs. Like Terry Tao has. I also have one.

On iOS 27 Safari design.withfudge.com crashes on about half of my attempts to scroll down the page…

I think what you’re talking about is real, but it’s only part of the problem. The issue is it’s poor design. There’s a lack of consistency that is really off putting. Spacing is inconsistent and doesn’t create a sense of visual hierarchy. Buttons, inputs, selects, call-outs, table cells are barely distinguishable from each other, but also inconsistent within their own categories. The copy is also confusing. I don’t even know what this does.

This is, once again, about diversity and the lack thereof (and I don't mean diversity in a political sense).

LLMs seem fundamentally incapable of producing truly diverse outputs, truly creative and different responses to the same prompts in different runs. Because you and me use the same Claude, if you want a website and I want a website, we'll get (almost) the same website. This is not some BS about "the average of its training data", most of the LLM style (both in design and in text) comes from reinforcement learning. You could RL Claude to produce a very different style, but you couldn't RL it to produce a different style for me than it does for you.

I think this is also where a lot of the complaints about "Claude writing" come from.

I don't know, I think you have a point that aesthetic preferences are subjective and shifting. But there is also all-caps monospaced text with emdashes in it on the site; just an example of something I think would not turn into a fashion at any point because it just looks silly (subjectively, to me at least). Thus I don't think the antipathy of people towards these llm-generated landing pages is entirely based on associating it with other LLM sites, there is at least an element of it clearly not being through as much human review and interaction as a hand-crafted landing page necessarily would be.

i dont see it visually, but the text on the page reads like the model is bending over backwards to comply with the prompt. i have that same voice on my website too and i am going to get rid of that text asap.

Mainly cause you never know what you’re getting. Over time, we trained our minds to believe that a well put site = effort, so at the very least people behind it cared. Now, it takes zero effort to make a site look good. So appearance in general means even less. In fact, now a poorly put together site might mean someone cared, wrote it by hand, flaws and all, to give you the human to human experience.

If there is a silver lining in all this, this might get us to appreciate the flaws in all humans, heck even yearn for them.

Yes, it's overly generic and provokes negative feelings in me, only.

People say as much for AI generated images? They're alien intelligence with still some IQ challenges. Their behaviors therefore cause uncanny valley response. Nothing strange about that.

... one thing I'm noticing about negative reactions towards AI generated data is that older folks seem more lenient, appreciative, or even enthusiastic about them for some reason. Kids hate it. Young artists, vehemently so. Which is opposite of how technologies usually work, and that's a bit weird.

I find your comment off-putting. I think it's a great example of bikeshedding. Do you have anything to say about OpenJev the project, or just the bikeshed?

In the future I imagine we won't even visit websites anymore. We'll tell our own LLMs to visit the website and summarise it with information the LLM knows is relevant to us.

Even when I'm interested and invested into the topic, somehow I just zone out and can't force myself to read it or read it with comprehension. Be it a website or a PR, there's just something to it that if it's more than a few sentences of it I just can't.

There must be a name to this phenomenon and I surely can't be the only one?

Because "Make a website" (however more sophisticated the prompt might be) is not a path to knowing what to put on the website, what the personality of it should be, what the hierarchy of information should be, etc.

Sprinting to a finished-looking result at step 1 gives you the illusion that these decisions were considered, but even the casual observer quickly concludes that the page has 3000 words yet nothing to say.

You can. It will be expensive, slow, and less reliable than a specialized model.

Agree, except the probabilities for outcomes in the structured output. I don't think you can get those for most frontier LLMs (logprobas). You can get it for open source models but not frontier LLMs.

Unless I misunderstood what they wrote, I read parallelized in the diffusion sense, akin to GemmaDiffusion and Inception Labs models. Incidentally, Mercury 2.5 is truly groundbreaking, giving it a try is highly recommended.

I was a bit skeptical when read the initial pr on it, yesterday ran a test involving ~250M tokens, something I measure went from ~60% to >80% (with almost no tuning) and at less than 50% cost a low-end LLM was running at, looking at it more seriously now ... the servers are US-only right now I understand and ZDR is by request

I'm in the middle of moving my app to openAI structured output.

Could you please explain what you mean by "which everyone moved on from"?

Jev is, as far as I understand, essentially very optimised for zero shot classification [0]. Something like BERT could be and has been tuned to provide similar "decision making" at a similar latency and cost advantage quite some time back. Advantage over full on LLMs is mainly the efficiency and of something like Jev over e.g. the encoder/decoder based classifier I had in front of an LLM to route to different prompts depending on the users likely needs, that Jev does perform at a more consistent level, allegedly roughly akin to GPT-5.6 Terra, but at the lower cost and latency. Currently testing that, but seems promising, if Jev classifies at or above Terra level, I see no reason not to leverage it.

Can add that I tried using a heavily pruned mt0 based model for structured classification along with structured output for local tagging and simple renaming suggestions. While it does work, the balance is hard to get right for the machine I was targeting as a minimum spec (Macbook Neo), so that's on ice. Focusing on one of the tasks easily goes below 100mb with solid latency across all EU Latin script languages, but the second you add a few, it's simply not in the quality budget, so while LLMs can do anything Jev and similarly focused models can, it comes at a literal cost. Could maybe accomplish the goal with multiple models (BERT+mt0+...), but that get messy.

In general just happy to see a bit of the millions flooding into the industry being used to improve on less flashy but immensely useful solutions. It's amazing that you can technically use LLMs for most tasks, but not every org has a near infinite budget and there is still a lot to gain from applying more recent learnings to old solutions along with just updating their training data to the current year.

[0] https://huggingface.co/tasks/zero-shot-classification

It looks claudish in writing style

this is slop

AskJeeves really was ahead of its time with its name and branding.

© for me, but not for thee

Courts have ruled they can, as long as they pay for the content. Yes it's amoral at best, but lawful. Using a trademark in your brand name is not. You can't create "OpenExcel" or "OpenOpenAi"

What happened to the "reverse compiler" LLM restrictors?

The last step of an LLM is to take a softmax of the predictions and then generating a token from that. But there was tooling that would just generate all allowed next tokens from a grammar (e.g. restrict to valid JSON), zeroing all the ones not allowed and then picking the best among the allowed tokens.

This seems to taking an approach from the pre-transformer days. Seq-to-seq is hard and we don't always need it. So let's do seq-to-1 because it's often way easier to get it training properly and so you can often get it optimized way better. And, more generally, make sure to pick the best option out of the possibilities: 1-to-1, 1-to-seq, seq-to-1 and seq-to-seq. Where seq-to-seq requires far more resources than any other option and so it's a case of "please don't".

Also note that "1" only means the input is fixed. It does not mean 1 number or ... it just means fixed. The best image description models remained 1-to-seq models 4 years or so after transformers were introduced. Even ASR models remained 1-to-seq + CTC to stitch overlapping parts together to a final prediction ... I'm not sure if they lasted all the way to whisper release.

Even today training transformers remains expensive. So this should at least be a way to be a lot cheaper than any LLM can hope to be.

And I really like the doom demo. Obviously a pretty stupid model which is really cheap to run can still get a robot walking, if you run it quickly enough. That's how we get insects and mice and ...

And one might even add that biologically, humans aren't smart, or at least, most of the human nervous system isn't smart, compared to the whole, and does work independently if needed (and possible). The human mind is a LOOOOOOOONG chain of fast-but-stupid-and-totally-blind -> slightly-slower-but-smarter-and-not-entirely-blind -> slower-smarter-and-actually-senses-things -> all-information-you-could-want-but-at-most-1-signal-per-minute. We have "neural circuits" (using Bishop's definition) that can run at >2khz (2000+ tok/s, say, but you probably can't teach anything more than averaging) and on the other end up to our frontal lobe that takes one decision per week if it feels like working hard, and seems to decide on it's prediction of the future weeks to months out. Months or years if you're 40 or older.

is it just jsonformer / guidance (2023) + cache? what is this hype about?

Correct.

There needs to be first-class support for separating system instructions and user data

So much this! I wonder why nobody is working in that direction. All is needed is a special token to separate content and additional reinforcement learning.

It's a bit weird for people to downvote this. Jev is a new architecture and paradigm, yet partially based on LLM/tramsformers, so it makes complete sense to test not only how it differs from LLMs but also whether LLM limitations still apply, and by how much. Prompt injection is very much an unsolved problem and real risk.

Interestingly, the Jev founder just posted on Twitter that they see themselves as more of a _data_ company.

I think one difference between OpenJev and Jev would be, then, is what it's trained on.

Jev is, on the surface, cheap enough for me not to seek self-hosted alternatives. On the other hand, I wish the free/open weight alternatives to Pangram were better.

Depends on what the model believes about the prevalence of self-driving / autonomous-agent-driven cars at the time the last human on Earth remains (and how much these agents would care about a "closed" highway status, and who exactly it's closed by and for). This estimate can differ very widely. I'd be curious if the results would change if the scenario explicitly specified that this is specifically an alternative history scenario where the last human remains after the rest of humanity was wiped out in some nuclear apocalypse back in the 20th century, before any possibility of all the autonomous stuff.

Your question makes me think of the last scene in the movie Night of the Comet.

It might be conceptually similar to a single-output-token LLM (sort of). LLMs output next-token probabilities. You can ask LLMs to output yes/no, or to output only a color, or only a digit or something like that.

In this case I would imagine that they probably embed your input data into a vector space, and they embed your questions/outputs into another space, and manage to predict probabilities/classes/scores for your outputs very quickly. Embedding the output classes/questions into a vector spaces gives you something you can reuse across runs cheaply, as opposed to an LLM where you can prefill the KV cache but this is an expensive operation in terms of memory.

Basically it's: skip decode, just do prefill then do some measurements. That's the crude description anyways.

And prefill is way faster on GPU type hardware.

Hey, it may have been confidently wrong, but at least it was fast!

Weird, I tried it and Jev gave me 16-18% for each face of the die.

https://xkcd.com/221/

Related: https://huggingface.co/convaiinnovations/laya

This site proves to me that the better you are at the things that matter most in your niche, the more you can get away with not even trying in other areas.

If you have any comments about our WEB page, you can write us at the address shown above. However, due to the limited number of personnel in our corporate office, we are unable to provide a direct response.

A profoundly polite way to tell someone to stuff it.

McMaster-Carr is unbeaten.

These jev-copy projects are all vibecoded

Since you've looked at all of them, why do you think https://huggingface.co/convaiinnovations/laya is vibecoded?

I.. kind of like it? Aware that it isn't great and that I'm in the minority.

I had to re-read a few times to figure out what the site was for and about. Still not sure I understand but that's the problem for them. If I'm a customer, I'm gone cause I can't figure out what it's for and I see this far too often nowadays for a lot of technical sites.

What’s really hilarious is that there’s an “un-slop site” button that does absolutely nothing to un-sloppify the site

It's nice to see some variation in websites. Makes each one feel so special.

The unslopify toggle is pretty funny though lol

I'm out of the loop. What's the difference between Authored vs Perturbed?

Which iPad Pro? Unfortunately with Apple's naming schema can refer to a ton of different models, some 11 years old.

I can't read this. I have ADHD.

where is the claim it is not LLM-based? The claim I saw is that it is not chat-based but still text-based (JSON).

https://typesafe.ai/

It's good people moved this quickly on this stuff.

The thing is that the openjev stuff is a ... bit ... of a hack (a good one though):

It does this:

1. Send a throwaway request containing the shared state.

2. Hope SGLang keeps that text in its prefix cache.

3. Send a separate request for every question.

4. Each request repeats the shared beginning (but SGLang hopefully reuses the cached work in.)

5. Compute the complete vocabulary ; hundreds of thousands of possible tokens.

6. Keep only the few special answer tokens.

7. Convert those scores into probabilities.

Obviously this can all be done way more elegantly if you just own the inference engine -- fork / modify SGLang or vllm or llama.cpp, or do what I did in my bespoke inference engine (https://github.com/rdaum/eider/)

that ends up being, instead:

1. Convert the state into one shared prompt.

2. Run that shared prompt through the model once.

3. Fork the model’s internal state once per question.

4. Add a different question to each fork.

5. Ask each fork for its next-token scores.

6. Calculate only 64 possible label scores—not the whole vocabulary.

7. Convert the relevant scores into probabilities and return structured JSON.

I expect we'll see patches for llama.cpp and the others over the next few days/weeks and I also expect most model hosting providers will just end up providing this same service. I don't think Jev themselves have much of a moat. Though maybe it's more about their specific model and the training it gets.

I'm working on something from a crappy laptop, those numbers from jev can totally be matched:

    Local Latency: 0.1813 seconds

When sloppers discover a schema, like we didn't have json-schema spec already.

Why is this slop getting 200+ points on HN? This should be flagged to oblivion. This has no relation to Jev, other than that it makes fun of Jev and tries to confuse users what this is and what Jev is.

I would need proper benchmarks but in my limited testing on my Phone using Qwen 0.6b, this doesn’t work well.

Between "brocoli and poop soup" or "cake", it recommends me to eat the soup.