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

I-have-ADHD: A skill to stop coding agents from burying the answer

github.com

Post reply on HN
Print view View original

341–350 of 390 posts

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#341

Post by prawn »

vzmax wrote: post #161
hysan wrote: post #120

Earlier quoted context omitted.

Relevant anecdata because I've burned many a Claude sessions on this. If you're using Claude Code, then it's in the harness. At the close of many sessions, I would start a meta conversation over why the LLM would consistently break certain rules. What it found when debugging itself is that some of the "contradicting" rules that I had were in fact, not from my rules. Instead, the instructions from its own harness had…

You can't trust it's account on why it did something, it does not "remember". It will just make up something plausible sounding.

I've had Claude completely invent a word, and then admit under questioning that it was because it seemed plausible.

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#342

Post by officialchicken »

dmos62 wrote: post #330
ryandrake wrote: post #74

The biggest "Claudism" that I have a hard time getting the LLM to stop doing is its insistence on talking about what it didn't do in addition to what it did. "I edited this.py and that.py but I did not edit README.md and I did not commit." or code comments like "This code invokes foo on bar and returns the results directly -- not through a callback." "This code returns true if the user clicked on a button -- not on t…

LLMs are trained to obey instructions, and they try their best to game the reinforcement learning by including reports of how they're obeying your instructions. Therefore, not talking about followed instructions is a sort of conflict for an LLM.

Trained to obey? More like instructed to obey in an observable manner.

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#344

Post by ffsm8 »

prawn wrote: post #341
vzmax wrote: post #161

Earlier quoted context omitted.

You can't trust it's account on why it did something, it does not "remember". It will just make up something plausible sounding.

I've had Claude completely invent a word, and then admit under questioning that it was because it seemed plausible.

[flagged]

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#345

Post by dmos62 »

officialchicken wrote: post #342
dmos62 wrote: post #330

Earlier quoted context omitted.

LLMs are trained to obey instructions, and they try their best to game the reinforcement learning by including reports of how they're obeying your instructions. Therefore, not talking about followed instructions is a sort of conflict for an LLM.

Trained to obey? More like instructed to obey in an observable manner.

I'm describing reinforcement learning.

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#346

Post by scrollaway »

ffsm8 wrote: post #344
prawn wrote: post #341

Earlier quoted context omitted.

I've had Claude completely invent a word, and then admit under questioning that it was because it seemed plausible.

[flagged]

Your HN comment is not useful. If you think it is, then you're suffering from HN psychosis.

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#347

Post by fortyseven »

idicjeifjwjd wrote: post #97
ryandrake wrote: post #74

The biggest "Claudism" that I have a hard time getting the LLM to stop doing is its insistence on talking about what it didn't do in addition to what it did. "I edited this.py and that.py but I did not edit README.md and I did not commit." or code comments like "This code invokes foo on bar and returns the results directly -- not through a callback." "This code returns true if the user clicked on a button -- not on t…

The thing I struggle the most with is getting it to stop referring to itself with personal pronouns. No Claude, you are not an “I” you are an “it”. You are a fucking tool, dammit. Tell me what you did without trying to assume personality; stop impersonating humans you steroidal autocorrect.

Who gives a shit? It can call itself the king of Norway for all I care as long as things get done correctly. If you're suggesting this is causing excessive token usage or something I think that's a bit exaggerated.

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#348

Post by lucas_t_a »

hysan wrote: post #120
ryandrake wrote: post #74

The biggest "Claudism" that I have a hard time getting the LLM to stop doing is its insistence on talking about what it didn't do in addition to what it did. "I edited this.py and that.py but I did not edit README.md and I did not commit." or code comments like "This code invokes foo on bar and returns the results directly -- not through a callback." "This code returns true if the user clicked on a button -- not on t…

Relevant anecdata because I've burned many a Claude sessions on this. If you're using Claude Code, then it's in the harness. At the close of many sessions, I would start a meta conversation over why the LLM would consistently break certain rules. What it found when debugging itself is that some of the "contradicting" rules that I had were in fact, not from my rules. Instead, the instructions from its own harness had…

you can always use /rewind to go back, always work thinking "can i shorten or clear my context now? and how should i do it?"

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#349

Post by zaphirplane »

ffsm8 wrote: post #344
prawn wrote: post #341

Earlier quoted context omitted.

I've had Claude completely invent a word, and then admit under questioning that it was because it seemed plausible.

[flagged]

100% sentences like i convinced it or made it admit are just weird But how else to describe the situation

Re: I-have-ADHD: A skill to stop coding agents from burying the answer

#350

Post by TomGarden »

bilalq wrote: post #239
TomGarden wrote: post #209

Earlier quoted context omitted.

Yeah agree. Someone on HN posed the theory that the dogfooding might have worked against them here - the theory is that internally they used Fable as conversational agent and Opus as subagents, leading to Opus falling deeper into a style mainly aimed at other LLMs. All conjecture of course, and yeah it's hard to imagine they would enjoy this prose internally

Fable can't write in a legible manner either.

I find it slightly more acceptable, but yeah agree, still bad.
Post reply on HN