Making o1, o3, and Sonnet 3.7 hallucinate for everyone
1–10 of 233 posts
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#2o3-mini or o3-mini-high?
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#3It’s not really hallucinating though, is it? It’s repeating a pattern in its training data, which is wrong but is presented in that training data (and by the author of this piece, but unintentionally) as being the solution to the problem. So this has more in common with an attack than a hallucination on the LLM’s part.
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#4This is interesting. If the models had enough actual code as training data, that forum post code should have very little weight, shouldn't it? Why do the LLMs prefer it?
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#5ChatGPT used to assure me that you can use JS dot notation to access elements in a Python dict. It also invented Redocly CLI flags that don't exist. Claude sometimes invents OpenAPI specification rules. Any time I ask anything remotely niche, LLMs are often bad.
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#6[flagged]
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#7ChatGPT used to assure me that you can use JS dot notation to access elements in a Python dict. It also invented Redocly CLI flags that don't exist. Claude sometimes invents OpenAPI specification rules. Any time I ask anything remotely niche, LLMs are often bad.
Any time I ask anything, LLMs are often bad.
inb4 you just aren't prompting correctly
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#8This is interesting. If the models had enough actual code as training data, that forum post code should have very little weight, shouldn't it? Why do the LLMs prefer it?
Probably because the coworker's question and the forum post are both questions that start with "How do I", so they're a good match. Actual code would be more likely to be preceded by... more code, not a question.
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#9ChatGPT used to assure me that you can use JS dot notation to access elements in a Python dict. It also invented Redocly CLI flags that don't exist. Claude sometimes invents OpenAPI specification rules. Any time I ask anything remotely niche, LLMs are often bad.
Yeah this is so common that I've already compiled a mental list of prompts to try against any new release. I haven't seen any improvement in quite a long while now, which confirms my belief that we've more or less hit the scaling wall for what the current approaches can provide. Everything new is just a microoptimization to game one of the benchmarks, but real world use has been identical or even worse for me.
Re: Making o1, o3, and Sonnet 3.7 hallucinate for everyone
#10It’s not really hallucinating though, is it? It’s repeating a pattern in its training data, which is wrong but is presented in that training data (and by the author of this piece, but unintentionally) as being the solution to the problem. So this has more in common with an attack than a hallucination on the LLM’s part.
Everything they do is hallucination, some of it ends up being useful and some of it not. The not useful stuff gets called confabulation or hallucination but it's no different from the useful stuff, generated the same exact way. It's all bullshit. Bullshit is actually useful though, when it's not so wrong that it steers people wrong.