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

Models Are Getting Dumber on Purpose

w4g1.dev

Post reply on HN
Print view View original

161–170 of 197 posts

Re: Models Are Getting Dumber on Purpose

#161

Post by brainless »

kennywinker wrote: post #4

Ideally what I'd like to see is pluggable knowledge bases. So if I'm e.g. coding a SwiftUI app for navigation, I'd take 9B of basic coding and reasoning, add 10B of swift/swiftUI, add 5B of GIS/geography knowledge and another 5B of frontend app design knowledge. My model doesn't need to know a single line of python. Then when I want to research electronics components, I grab a 15B model of agentic research techniques…

This is how my experiments go. And I am sure there are popular agents that do this. How I am trying is to create "Rust Engineer", "Typescript Engineer" or even "Rust Diesel Engineer". I have not tried fine-tuning. I focus on a small model, usually Qwen3.5 9B. I take a bunch of open source repositories and build a KG on it. A small model should be able to enrich your prompt and add technical context. The final, enriched prompt goes to the capable model.

Re: Models Are Getting Dumber on Purpose

#163

Post by skissane »

kennywinker wrote: post #13
msdz wrote: post #9

Great article, even if it will be interesting to see whether things continue to develop in such a direction or not. > There's a version of this future where the model card stops listing a knowledge cutoff at all, because what's left in the weights goes stale on a scale of years instead of weeks. Future? Even just recently I’ve read of two approaches to this problem: Cactus have come up with Needle [0][1], which is th…

That kind of setup is super dependent on a search engine, and search keeps getting worse.

> search keeps getting worse

For a lot of use cases, you don't need a general purpose search engine – a search over a curated knowledge base works even better.

There are plenty of freely available data sets you can use, depending on the application; plus in many cases you will want to use internal-only knowledge bases containing non-public information (e.g. documentation for a corporation's internal systems and procedures). There are also many paid subscription domain-specific knowledge services available.

Re: Models Are Getting Dumber on Purpose

#164

Post by Izkata »

kennywinker wrote: post #90
hoppp wrote: post #63

Earlier quoted context omitted.

That's still not hearing. Hearing has volume, direction, pitch, it's spacial processing etc

Laurel / yanny has entered the chat. Hearing is also psycho, not just acoustic.

Brainstorm / Green Needle: https://www.youtube.com/watch?v=1okD66RmktA

I can change which one I hear on a whim, so much that I can even get Brain Needle and Green Storm out of this video.

Re: Models Are Getting Dumber on Purpose

#165

Post by margalabargala »

glenstein wrote: post #159
kentonv wrote: post #99

Earlier quoted context omitted.

Tangent: This is often true of humans as well. We often make a decision based on a gut feeling, and then backfill a logical reason supporting our feeling, without even realizing we're doing it -- rationalization.

Tangent on the tangent: I think that's true in a minority of cases and in a majority of AI cases. Though in principle I think it should be possible for an LLM to have access to and faithfully represent its own reasoning.

On the contrary, I would argue that it's true in a totality of AI cases.

To your point, I agree that nominally there should be a way to give conceptual names to paths of weights, and when answering a question, notice which weights were and were not applied and retrospect on that.

That's not what reasoning traces as they currently exist are, though.

Re: Models Are Getting Dumber on Purpose

#166

Post by ripley12 »

lowsong wrote: post #69

> This mostly solves hallucination The author is factually incorrect here. Moving information out of the model weights and into the input of the model's context window in no way ensures that the model will accurately output content that was input from the context. This is true even when RAG is used to input exactly the correct data.

It does not guarantee accuracy, but it helps a lot in my experience; I find that models tend to hallucinate a lot more about things that aren't in their context window.

Re: Models Are Getting Dumber on Purpose

#167

Post by efilife »

angry_octet wrote: post #153
kentonv wrote: post #99

Earlier quoted context omitted.

Tangent: This is often true of humans as well. We often make a decision based on a gut feeling, and then backfill a logical reason supporting our feeling, without even realizing we're doing it -- rationalization.

And we all know some people that rationalise poor choices and misbehavior, hide their mistakes, etc, to an unacceptable degree. Sometimes the individual knows they are rationalising but continues anyway, other times they seem incapable of seeing that. When you ask people who are rationalising poor behaviour about the scenario, but it is someone else doing it, they may arrive at a better answer. Can we use multiple LL…

isn't this happening already? There's the concept called "thinking" where the models talks with itself before giving you the final answer

Re: Models Are Getting Dumber on Purpose

#168

Post by kennywinker »

dools wrote: post #125
kennywinker wrote: post #4

Ideally what I'd like to see is pluggable knowledge bases. So if I'm e.g. coding a SwiftUI app for navigation, I'd take 9B of basic coding and reasoning, add 10B of swift/swiftUI, add 5B of GIS/geography knowledge and another 5B of frontend app design knowledge. My model doesn't need to know a single line of python. Then when I want to research electronics components, I grab a 15B model of agentic research techniques…

This just means you have to describe what you want in Swift or whatever. If it doesn’t have the language then it doesn’t have the capability to transform intent into code.

In my hypothetical, 9B of coding and reasoning includes enough language to communicate ideas to the model.

Re: Models Are Getting Dumber on Purpose

#169

Post by teravor »

twalla wrote: post #113
chopete3 wrote: post #67

Earlier quoted context omitted.

This is so right. We training Whisper Large model on 20,000 audio samples specific to a domain and it ended up reducing the ASR by 5% while improving WER of the finetuned domain by 0.5%. Instead we ended up with no finetuning. We give audio snippet to 2 AsR models, take 3 best transcriptions and ask the LLm to pick the best based on the context. That produced significantly higher accuracy in how an agent understands…

Can you go into more detail on this? I’ve been putting together the data pipeline for an ASR fine tune but your approach sounds more interesting.

basically he is feeding the same input to multiple models, taking their outputs and dumping it into an LLM to sort out what the actual transcription probably is. expensive but effective.

Re: Models Are Getting Dumber on Purpose

#170

Post by dosisking »

kelnos wrote: post #112
kennywinker wrote: post #4

Ideally what I'd like to see is pluggable knowledge bases. So if I'm e.g. coding a SwiftUI app for navigation, I'd take 9B of basic coding and reasoning, add 10B of swift/swiftUI, add 5B of GIS/geography knowledge and another 5B of frontend app design knowledge. My model doesn't need to know a single line of python. Then when I want to research electronics components, I grab a 15B model of agentic research techniques…

> I don't want general purpose models. They try to be everything to everyone. I think the vast majority of people do want general purpose models. They want to be able to ask it any question, or ask it to perform any task, and for it to do a decent job at it. I agree that it's really hard (maybe even impossible) to build something that's everything for everyone. But your average (or even above-average) LLM user doesn'…

> I think the vast majority of people do want general purpose models. They want to be able to ask it any question, or ask it to perform any task, and for it to do a decent job at it.

The vast majority of people listen to the music of Ed Sheeran and think that it is good.

Post reply on HN