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

People are just as bad as my LLMs

wilsoniumite.com

Post reply on HN
Print view View original

51–60 of 173 posts

Re: People are just as bad as my LLMs

#51

Post by dkenyser »

mplewis wrote: post #45
smallnix wrote: post #20

Is my understanding wrong that LLMs are trained to emulate observed human behavior in their training data? From that follows that LLMs fit to produce all kinds of human biases. Like preferring the first choice out of many, and the last our of many (primacy biases). Funnily the LLM might replicate the biases slightly wrong and by doing so produce new derived biases.

LLMs don't emulate human behavior. They spit out chunks of words in an order that parrots some of their training data.

Correct me if I'm wrong, but I feel like we're splitting hairs here.

> spits out chunks of words in an order that parrots some of their training data.

So, if the data was created by humans then how is that different from "emulating human behavior?"

Genuinely curious as this is my rough interpretation as well.

Re: People are just as bad as my LLMs

#52

Post by sedatk »

robwwilliams wrote: post #21
tehsauce wrote: post #3

There has been some good research published on this topic of how RLHF, ie aligning to human preferences easily introduces mode collapse and bias into models. For example, with a prompt like: "Choose a random number", the base pretrained model can give relatively random answers, but after fine tuning to produce responses humans like, they become very biased towards responding with numbers like "7" or "42".

I assume 42 is a joke from deep history and The Hitchhiker’s Guide. Pretty amusing to read the Wikipedia entry: https://en.wikipedia.org/wiki/42_(number)

Douglas Adams picked 42 randomly though. :)

Re: People are just as bad as my LLMs

#53

Post by nthingtohide »

smallnix wrote: post #20

Is my understanding wrong that LLMs are trained to emulate observed human behavior in their training data? From that follows that LLMs fit to produce all kinds of human biases. Like preferring the first choice out of many, and the last our of many (primacy biases). Funnily the LLM might replicate the biases slightly wrong and by doing so produce new derived biases.

Not only that if future AI distrusts humanity it is because history, literature and fiction is full of such scenarios and AI will learn those patterns and associated emotions from those texts. Humanity together will be responsible for creating a monster (if that scenario happens).

Re: People are just as bad as my LLMs

#54

Post by sumeno »

mplewis wrote: post #45
smallnix wrote: post #20

Is my understanding wrong that LLMs are trained to emulate observed human behavior in their training data? From that follows that LLMs fit to produce all kinds of human biases. Like preferring the first choice out of many, and the last our of many (primacy biases). Funnily the LLM might replicate the biases slightly wrong and by doing so produce new derived biases.

LLMs don't emulate human behavior. They spit out chunks of words in an order that parrots some of their training data.

This is a more pedantic and meme-y way of saying the same thing.

Re: People are just as bad as my LLMs

#55

Post by geonineties »

mdp2021 wrote: post #46
consumer451 wrote: post #37

Earlier quoted context omitted.

Maybe misanthropic?

Nope. Misanthropic is when some people dislike other people. Racist is when some people attribute some dubious quality to all people in some category.

(dropped the snark) Racist means grouping according to race, or potentially geographic origins. The word for what you're describing is probably closest to discriminatory, or prejudicial.

However, misanthropic is probably more correct as the paper applies to all people negatively.

Re: People are just as bad as my LLMs

#56

Post by yashap »

Jensson wrote: post #29
d4mi3n wrote: post #11

Earlier quoted context omitted.

My guess is that we bias towards numbers with cultural or personal significance. 7 is lucky in western cultures and is religiously significant (see https://en.wikipedia.org/wiki/7#Culture ). 42 is culturally significant in science fiction, though that's a lot more recent. There are probably other examples, but I imagine the mean converges on numbers with multiple cultural touchpoints.

I have never heard of 7 being a lucky number in western culture and your link doesn't support that. 3 is a lucky number, 13 is an unlucky number, 7 is nothing to me. So I don't think its that, 7 is still a very common "random number" here even though there is no special cultural significance to it.

Hmm really? Even on the Wikipedia page for 7 (https://en.m.wikipedia.org/wiki/7), one of the first things it says is “7 is often considered lucky in Western culture and is often seen as highly symbolic.” And FWIW you can see the Wikipedia edit history, that isn’t a recent edit, nobody here is messing with it :)

“Lucky Number 7” is a common phrase, there was even a popular movie that played on this, “Lucky Number Slevin” (https://m.imdb.com/title/tt0425210/). It’s one of the first numbers I’d think of as a “lucky number.”

Re: People are just as bad as my LLMs

#57

Post by robertlagrant »

sedatk wrote: post #52
robwwilliams wrote: post #21

Earlier quoted context omitted.

I assume 42 is a joke from deep history and The Hitchhiker’s Guide. Pretty amusing to read the Wikipedia entry: https://en.wikipedia.org/wiki/42_(number)

Douglas Adams picked 42 randomly though. :)

Not at all. It was derived mathematically from the Question: What do you get if you multiply six by nine?

Re: People are just as bad as my LLMs

#58

Post by p1necone »

aidos wrote: post #8
tehsauce wrote: post #3

There has been some good research published on this topic of how RLHF, ie aligning to human preferences easily introduces mode collapse and bias into models. For example, with a prompt like: "Choose a random number", the base pretrained model can give relatively random answers, but after fine tuning to produce responses humans like, they become very biased towards responding with numbers like "7" or "42".

Why is that ? Whenever I’m giving examples I almost always use 7, something ending in a 7 or something in the 70s

1 and 10 are on the boundary, that's not random so those are out.

5 is exactly halfway, that's not random enough either, that's out.

2, 4, 6, 8 are even and even numbers are round and friendly and comfortable, those are out too.

9 feels too close to the boundary, it's out.

That leaves 3 and 7, and 7 is more than 3 so it's got more room for randomness in it right?

Therefore 7 is the most random number between 1 and 10.

Re: People are just as bad as my LLMs

#59

Post by MyOutfitIsVague »

educasean wrote: post #47
mplewis wrote: post #45

Earlier quoted context omitted.

LLMs don't emulate human behavior. They spit out chunks of words in an order that parrots some of their training data.

Is this just pedantry or is there some insight to be gleaned by the distinction you made?

I can only assume that either they are trying to point out that words aren't behavior, and mimicking human writing isn't the same thing as mimicking human behavior, or it's some pot-shot at the capabilities of LLMs.
Post reply on HN