Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Gemini 3 Pro: the frontier of vision AI

blog.google

Post reply on HN
Print view View original

211–220 of 309 posts

Re: Gemini 3 Pro: the frontier of vision AI

#211

Post by runarberg »

adastra22 wrote: post #207
runarberg wrote: post #184

Earlier quoted context omitted.

Hallucination are not generalization beyond the training data but interpolations gone wrong. LLMs are in fact good at generalizing beyond their training set, if they wouldn’t generalize at all we would call that over-fitting, and that is not good either. What we are talking about here is simply a bias and I suspect biases like these are simply a limitation of the technology. Some of them we can get rid of, but—like a…

What, may I ask, is the difference between "generalization" and "interpolation"? As far as I can tell, the two are exactly the same thing. In which case the only way I can read your point is that hallucinations are specifically incorrect generalizations. In which case, sure if that's how you want to define it. I don't think it's a very useful definition though, nor one that is universally agreed upon. I would say a h…

Interpolation and generalization are two completely different constructs. Interpolation is when you have two data points and make a best guess where a hypothetical third point should fit between them. Generalization is when you have a distribution which describes a particular sample, and you apply it with some transformation (e.g. a margin of error, a confidence interval, p-value, etc.) to a population the sample is representative of.

Interpolation is a much narrower construct then generalization. LLMs are fundamentally much closer to curve fitting (where interpolation is king) then they are to hypothesis testing (where samples are used to describe populations), though they certainly do something akin to the latter to.

The bias I am talking about is not a bias in the training data, but bias in the curve fitting, probably because of mal-adjusted weights, parameters, etc. And since there are billions of them, I am very skeptical they can all be adjusted correctly.

Re: Gemini 3 Pro: the frontier of vision AI

#212

Post by raw_anon_1111 »

macNchz wrote: post #144
Rover222 wrote: post #64

Earlier quoted context omitted.

I just tried to get Gemini to produce an image of a dog with 5 legs to test this out, and it really struggled with that. It either made a normal dog, or turned the tail into a weird appendage. Then I asked both Gemini and Grok to count the legs, both kept saying 4. Gemini just refused to consider it was actually wrong. Grok seemed to have an existential crisis when I told it it was wrong, becoming convinced that I ha…

An interesting test in this vein that I read about in a comment on here is generating a 13 hour clock—I tried just about every prompting trick and clever strategy I could come up with across many image models with no success. I think there's so much training data of 12 hour clocks that just clobbers the instructions entirely. It'll make a regular clock that skips from 11 to 13, or a regular clock with a plaque saying…

It was ugly. But I got ChatGPT to cheat and do it

https://chatgpt.com/share/6933c848-a254-8010-adb5-8f736bdc70...

This is the SVG it created.

https://imgur.com/a/LLpw8YK

Re: Gemini 3 Pro: the frontier of vision AI

#213

Post by wasmainiac »

Workaccount2 wrote: post #48

Well It is the first model to get partial-credit on an LLM image test I have. Which is counting the legs of a dog. Specifically, a dog with 5 legs. This is a wild test, because LLMs get really pushy and insistent that the dog only has 4 legs. In fact GPT5 wrote an edge detection script to see where "golden dog feet" met "bright green grass" to prove to me that there were only 4 legs. The script found 5, and GPT-5 the…

Try the 6 legged Eni dog next!

Re: Gemini 3 Pro: the frontier of vision AI

#214

Post by nearbuy »

Rover222 wrote: post #64
Workaccount2 wrote: post #48

Well It is the first model to get partial-credit on an LLM image test I have. Which is counting the legs of a dog. Specifically, a dog with 5 legs. This is a wild test, because LLMs get really pushy and insistent that the dog only has 4 legs. In fact GPT5 wrote an edge detection script to see where "golden dog feet" met "bright green grass" to prove to me that there were only 4 legs. The script found 5, and GPT-5 the…

I just tried to get Gemini to produce an image of a dog with 5 legs to test this out, and it really struggled with that. It either made a normal dog, or turned the tail into a weird appendage. Then I asked both Gemini and Grok to count the legs, both kept saying 4. Gemini just refused to consider it was actually wrong. Grok seemed to have an existential crisis when I told it it was wrong, becoming convinced that I ha…

My guess is the part of its neural network that parses the image into a higher level internal representation really is seeing the dog as having four legs, and intelligence and reasoning in the rest of the network isn't going to undo that. It's like asking people whether "the dress" is blue/black or white/gold: people will just insist on what they see, even if what they're seeing is wrong.

Re: Gemini 3 Pro: the frontier of vision AI

#215

Post by Borealid »

vunderba wrote: post #77
Workaccount2 wrote: post #48

Well It is the first model to get partial-credit on an LLM image test I have. Which is counting the legs of a dog. Specifically, a dog with 5 legs. This is a wild test, because LLMs get really pushy and insistent that the dog only has 4 legs. In fact GPT5 wrote an edge detection script to see where "golden dog feet" met "bright green grass" to prove to me that there were only 4 legs. The script found 5, and GPT-5 the…

Anything that needs to overcome concepts which are disproportionately represented in the training data is going to give these models a hard time. Try generating: - A spider missing one leg - A 9-pointed star - A 5-leaf clover - A man with six fingers on his left hand and four fingers on his right You'll be lucky to get a 25% success rate. The last one is particularly ironic given how much work went into FIXING the ol…

You didn't correct the hands being backwards? It gave you a man with six fingers on his right hand, and four fingers on his left.

Re: Gemini 3 Pro: the frontier of vision AI

#216

Post by nearbuy »

JamesSwift wrote: post #205
nl wrote: post #204

Earlier quoted context omitted.

Programs can solve mazes and LLMs can program. That's a different thing completely.

That just seems like an arbitrary limitation. Its like asking someone to do answer a math calculation but "no thinking allowed". Like, I guess we can gauge if a model just _knows all knowable things in the universe_ using that method... but anything of any value that you are gauging in terms of 'intelligence', is going to actually be validating their ability to go "outside the scope" of what they actually are (an aut…

It depends whether you're asking it to solve a maze because you just need something that can solve mazes, or if you're trying to learn something about the model's abilities in different domains. If it can't solve a maze by inspection instead of writing a program to solve it, that tells you something about its visual reasoning abilities, and that can help you predict how they'll perform on other visual reasoning tasks that aren't easy to solve with code.

Re: Gemini 3 Pro: the frontier of vision AI

#217

Post by 8n4vidtmkvmk »

reducesuffering wrote: post #81
knollimar wrote: post #33

I do some electrical drafting work for construction and throw basic tasks at LLMs. I gave it a shitty harness and it almost 1 shotted laying out outlets in a room based on a shitty pdf. I think if I gave it better control it could do a huge portion of my coworkers jobs very soon

"AI could never replace the creativity of a human" "Ok, I guess it could wipe out the economic demand for digital art, but it could never do all the autonomous tasks of a project manager" "Ok, I guess it could automate most of that away but there will always be a need for a human engineer to steer it and deal with the nuances of code" "Ok, well it could never automate blue collar work, how is it gonna wrench a pipe i…

I still haven't witnessed a serious attempt at passing the Turing test. Are we just assuming its been beaten, or have people tried?

Like if you put someone in an online chat and ask them to identify if the person they're talking to is a bot or not, you're telling me your average joe honestly can't tell?

A blog post or a random HN comment, sure, it can be hard to tell, but if you allow some back and forth.. i think we can still sniff out the AIs.

Re: Gemini 3 Pro: the frontier of vision AI

#218

Post by seanmcdirmid »

nearbuy wrote: post #216
JamesSwift wrote: post #205

Earlier quoted context omitted.

That just seems like an arbitrary limitation. Its like asking someone to do answer a math calculation but "no thinking allowed". Like, I guess we can gauge if a model just _knows all knowable things in the universe_ using that method... but anything of any value that you are gauging in terms of 'intelligence', is going to actually be validating their ability to go "outside the scope" of what they actually are (an aut…

It depends whether you're asking it to solve a maze because you just need something that can solve mazes, or if you're trying to learn something about the model's abilities in different domains. If it can't solve a maze by inspection instead of writing a program to solve it, that tells you something about its visual reasoning abilities, and that can help you predict how they'll perform on other visual reasoning tasks…

You could actually add mazes and paths through them to the training corpus, or make a model for just solving mazes. I wonder how effective it would be, I’m sure someone has tried it. I doubt it would generalize enough to give the AI new visual reasoning capabilities beyond just solving mazes.

Re: Gemini 3 Pro: the frontier of vision AI

#219

Post by JamesSwift »

nearbuy wrote: post #216
JamesSwift wrote: post #205

Earlier quoted context omitted.

That just seems like an arbitrary limitation. Its like asking someone to do answer a math calculation but "no thinking allowed". Like, I guess we can gauge if a model just _knows all knowable things in the universe_ using that method... but anything of any value that you are gauging in terms of 'intelligence', is going to actually be validating their ability to go "outside the scope" of what they actually are (an aut…

It depends whether you're asking it to solve a maze because you just need something that can solve mazes, or if you're trying to learn something about the model's abilities in different domains. If it can't solve a maze by inspection instead of writing a program to solve it, that tells you something about its visual reasoning abilities, and that can help you predict how they'll perform on other visual reasoning tasks…

Again, think about how the models work. They generate text sequentially. Think about how you solve the maze in your mind. Do you draw a line direct to the finish? No, it would be impossible to know what the path was until you had done it. But at that point you have now backtracked several times. So, what could a model _possibly_ be able to do for this puzzle which is "fair game" as a valid solution, other than magically know an answer by pulling it out of thin air?

Re: Gemini 3 Pro: the frontier of vision AI

#220

Post by adastra22 »

runarberg wrote: post #211
adastra22 wrote: post #207

Earlier quoted context omitted.

What, may I ask, is the difference between "generalization" and "interpolation"? As far as I can tell, the two are exactly the same thing. In which case the only way I can read your point is that hallucinations are specifically incorrect generalizations. In which case, sure if that's how you want to define it. I don't think it's a very useful definition though, nor one that is universally agreed upon. I would say a h…

Interpolation and generalization are two completely different constructs. Interpolation is when you have two data points and make a best guess where a hypothetical third point should fit between them. Generalization is when you have a distribution which describes a particular sample, and you apply it with some transformation (e.g. a margin of error, a confidence interval, p-value, etc.) to a population the sample is…

I assumed you were speaking by analogy, as LLMs do not work by interpolation, or anything resembling that. Diffusion models, maybe you can make that argument. But GPT-derived inference is fundamentally different. It works via model building and next token prediction, which is not interpolative.

As for bias, I don’t see the distinction you are making. Biases in the training data produce biases in the weights. That’s where the biases come from: over-fitting (or sometimes, correct fitting) of the training data. You don’t end up with biases at random.

Post reply on HN