Hacker News Reader

Comment by hailwren | Hacker News Reader

It has always seemed to me that they're hacking for dopamine response in moderately interested data labelers.

Even when I add multiple prompts into the claude.md file not to be so sycophant sounding and just be blunt, it's responses are full of "the reason it lands...", "that's not X, it's Y" "Your understanding of X — it's better than most people's" or "you already own the right question...".

I don't like that I like it.

The most helpful instructions I've found that curb this: "Do not use superlatives. Do not use persuasive writing style."

I have other more specific ones to avoid talking about things that it's not doing, but those two sentences have covered a lot of ground for me when working w/ Opus models.

I have had success in rooting these out by using the correct linguistic terminology for each. Negative parallelisms, tricolons/polycolons, etc. I haven't come up with the proper terminology for all of them.

Interesting. I've found using the keyword "accretion" very useful for LLM code review.

Yes! The Claudisms do seem to have this slightly uncanny clickbaity feel to them.

I always thought it could be because volume-wise, most English prose is probably marketing copy and actual clickbait; so when you train on the entire Internet, you get a troll adept at writing ads. Then people ask AdBot2000 to write a novel and are upset it reads like the next iPhone launch site.

No, there's no reason chatbot behavior would have anything to do with frequency of text in pretraining.

Nah, I think this is a common misunderstanding of how LLMs work, where people think that they mimic the pre-training data. Stylistically everything you see is an artifact of post-training, which is from reinforcement learning not from absorbing mass amounts of text. At some point a person or more recently a bot gave a thumbs up to an A/B tested response including em-dashes and claudisms galore.

So question then, why is it so hard to make an ai that doesn’t do these things? And why do Claude and ChatGPT have the same -isms? They’re both doing the same a/b post training with the same decisions?

It would require changing humans first.

You don't blame the puddle for taking the shape of the hole.

Yes. This completely explains sycophancy at least.

There's layers, some of token selection is fingerprinting https://github.com/google-deepmind/synthid-text

Yeah, but I understand that fingerprinting is essentially a pseudorandom overlay onto a pseudorandom base signal. And unless you have access to both the random number generators and the weights, I don't think you can detect it?

So "fingerprinting" operates on a totally different and basically invisible level, as opposed to the obvious stylistic patterns that the average programmer can identify in about 2 sentences.

Eh we can detect opumism and gptisms our brain are very good at pattern recognition even if subconscious

> Stylistically everything you see is an artifact of post-training,

It is still not exactly clear if it is true or not. Unless we have base "pt" snaphot of Claude we can't say one way or another. I've played a bit with base models of Nemo, Gemma etc and they all had tics, not much different from RLHFed instruct versions.

It's more likely that this is from the training data if they're being trained on reams of Internet stuff.

Isn’t most of the internet slop by now? Self-reinforcing feedback loop.

See: upvotes here

To me it has a writerly New Yorker vibe to it, as in the magazine which reads as “polished” and probably performs well in RL but is totally exhausting to read in long sessions and completely inappropriate for coding where precision is paramount above all. In writing terms its called purple prose.

https://en.wikipedia.org/wiki/Purple_prose

The New Yorker may be pretentious but it's generally not unreadable like Opus.

Claude is unreadable and sometimes pretentious.-

You’re more right than you probably realize!

It's not clickbait, it's automated empathy!

from a few days ago https://news.ycombinator.com/item?id=49388752

Interesting! My impression was that this was an artifact of RLVR where this slightly preferred writing style got amplified to the nth degree. It's probably some mix.

Given how frequently this kind of punchy-but-vacuous slop gets voted onto the hn front page, the hacking seems to be working.

I assumed they just raw dogged the internet and if you do that, you see way more of that garbage than anything else. It's just that most of us have visually/mentally ignored all of that either via spam filters or just, you know, scrolled passed it.

Reply on news.ycombinator.com