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

Gemma 3n preview: Mobile-first AI

developers.googleblog.com

Post reply on HN
Print view View original

161–170 of 179 posts

Re: Gemma 3n preview: Mobile-first AI

#161

Post by tgtweak »

rvnx wrote: post #130

Absolute shit. Comparing it to Sonnet 3.7 is an insult. # Is Eiffel Tower or a soccer ball bigger ? > A soccer ball is bigger than the Eiffel Tower! Here's a breakdown: > Eiffel Tower: Approximately 330 meters (1,083 feet) tall. > Soccer Ball: A standard soccer ball has a circumference of about 68-70 cm (27-28 inches). > While the Eiffel Tower is very tall, its base is relatively small compared to its height. A socce…

Took a picture of a bag of chips and it said it was seasoning... I think, for the size, it's OK - but not really there. I'm not sure how they got ELO anywhere near Claude or Gemini... those models are leagues ahead in terms of one-shot accuracy.

Re: Gemma 3n preview: Mobile-first AI

#162

Post by anonzzzies »

disgruntledphd2 wrote: post #143
anonzzzies wrote: post #138

Earlier quoted context omitted.

>anything redolent of interaction with intelligence compared to what you are used to right? I know it's elitist but most people My circle of people I talk with during the day has changed since I took on more charity which consists of fixing up old laptops and installing Ubuntu on them; I get them for free from everyone and I give them to people who cannot afford, including some lessons and remote support (which is ea…

> but most people This is incorrect, IQ tests are normally scaled such that average intelligence is 100, and such that they are approximately normally distributed so that most people will be somewhere between 85-115 (66% on average).

Yep and those people can never 'win' against current llms, let alone future ones. Outside motorcontrol which I specifically excluded.

85 is special housing where I live... LLMs are far beyond that now.

Re: Gemma 3n preview: Mobile-first AI

#163

Post by lostmsu »

osanseviero wrote: post #105
jdiff wrote: post #68

Earlier quoted context omitted.

There's a 7B mentioned in the chat arena ELO graph, I don't see any other references to it though.

Hi! The model is 8B if you also load the vision and audio components. We just used the text model in LMArena.

Are vision and audio components available yet?

Re: Gemma 3n preview: Mobile-first AI

#164

Post by numpad0 »

bilbo0s wrote: post #148
numpad0 wrote: post #70

Earlier quoted context omitted.

They earn from in-app purchases too!

Not anymore! Only half joking. I really do think the majority of that revenue will be going away.

Oh no. It just could. The App Store is slowly regressing into a gambling area and there are obviously people in power that don't like it. I think if it were to go, it'll take Android sideloading with it and we'll be miles closer to some computing doomsday scenario. Oh no.

Re: Gemma 3n preview: Mobile-first AI

#165

Post by QuadmasterXLII »

onlyrealcuzzo wrote: post #10
krackers wrote: post #8

Earlier quoted context omitted.

Thank you, that helped a bit, although it's still not clear what exactly those parameters _are_. "Per-Layer Embedding (PLE) parameters that are used during model execution to create data that enhances the performance of each model layer." is too vague, and I can't find any other reference to "per-layer embedding parameters" in literature.

A layer is a transformer block / layer (basically the building block of the modern LLM architectures) - maybe Gemini can help you: https://gemini.google.com/share/cc58a7c6089e

lmao

Re: Gemma 3n preview: Mobile-first AI

#166

Post by selcuka »

disgruntledphd2 wrote: post #143
anonzzzies wrote: post #138

Earlier quoted context omitted.

>anything redolent of interaction with intelligence compared to what you are used to right? I know it's elitist but most people My circle of people I talk with during the day has changed since I took on more charity which consists of fixing up old laptops and installing Ubuntu on them; I get them for free from everyone and I give them to people who cannot afford, including some lessons and remote support (which is ea…

> but most people This is incorrect, IQ tests are normally scaled such that average intelligence is 100, and such that they are approximately normally distributed so that most people will be somewhere between 85-115 (66% on average).

> but most people > average intelligence is 100

You both are saying the same thing.

IQ is defined such that both average and mean would be equal 100. The combination of sub-100 and exactly-100 would be more people than above-100, hence "most people <=100 iq".

Re: Gemma 3n preview: Mobile-first AI

#167

Post by tptacek »

selcuka wrote: post #166
disgruntledphd2 wrote: post #143

Earlier quoted context omitted.

> but most people This is incorrect, IQ tests are normally scaled such that average intelligence is 100, and such that they are approximately normally distributed so that most people will be somewhere between 85-115 (66% on average).

> but most people > average intelligence is 100 You both are saying the same thing. IQ is defined such that both average and mean would be equal 100. The combination of sub-100 and exactly-100 would be more people than above-100, hence "most people <=100 iq".

Both average and mean, you say.

Re: Gemma 3n preview: Mobile-first AI

#168

Post by disgruntledphd2 »

anonzzzies wrote: post #162
disgruntledphd2 wrote: post #143

Earlier quoted context omitted.

> but most people This is incorrect, IQ tests are normally scaled such that average intelligence is 100, and such that they are approximately normally distributed so that most people will be somewhere between 85-115 (66% on average).

Yep and those people can never 'win' against current llms, let alone future ones. Outside motorcontrol which I specifically excluded. 85 is special housing where I live... LLMs are far beyond that now.

I'm not convinced this is true. I suspect that they'd mostly fail a version of Ravens matrices that didn't appear in the training set.

Re: Gemma 3n preview: Mobile-first AI

#169

Post by selcuka »

tptacek wrote: post #167
selcuka wrote: post #166

Earlier quoted context omitted.

> but most people > average intelligence is 100 You both are saying the same thing. IQ is defined such that both average and mean would be equal 100. The combination of sub-100 and exactly-100 would be more people than above-100, hence "most people <=100 iq".

Both average and mean, you say.

Oops, I meant both median and mean.

Re: Gemma 3n preview: Mobile-first AI

#170

Post by otabdeveloper4 »

croes wrote: post #140
Zambyte wrote: post #136

Earlier quoted context omitted.

People say nonsense all the time. LLMs also don't have this issue all the time. They are also often right instead of saying things like this. If this reply was meant to be a demonstration of LLMs not having human level understanding and reasoning, I'm not convinced.

Not people, some people. But a LLM is sometimes the genius and sometimes the idiot. That doesn’t happen often if you always talk to the same person

"Doing 'A' is wrong, to fix it you need to do 'A'" is not an idiotic statement. It's complete non-sense that no human has ever uttered.

But for LLMs these kinds of illogical statements are common.

LLMs don't think; what they do cannot be in good faith called "thinking" by any definition.

Post reply on HN