A useful pattern for LLM classification at scale is to let it hallucinate plausible, fake entities. Then resolve to real ones.
Hallucination by proxy in LLM-assisted differential diagnosis
Current evidence suggests that LLM assistance could augment the diagnostic accuracy of clinicians. However, these systems are black boxes, susceptible to hallucinations, and project a potentially...
Hallucination by proxy in LLM-assisted differential diagnosis
Current evidence suggests that LLM assistance could augment the diagnostic accuracy of clinicians. However, these systems are black boxes, susceptible to hallucinations, and project a potentially...
Why do LLMs make stuff up? New research peers under the hood.
Claude's faulty "known entity" neurons sometimes override its "don't answer" circuitry.
LLMs are bullshitters. But that doesn't mean they're not useful.
Note: This is a personal essay by Matt Ranger, Kagi's head of ML In 1986, Harry Frankfurt wrote On Bullshit. He differentiates a lying from bullshitting: Lying means you have a concept of what is true,...
Artificial
An LLM is a computer program. We should talk about it like a computer program.
Artificial
An LLM is a computer program. We should talk about it like a computer program.
My AI Skeptic Friends Are All Nuts
My smartest friends have bananas arguments about LLM coding.
My AI Skeptic Friends Are All Nuts
My smartest friends have bananas arguments about LLM coding.
What Happens, Exactly, When a Person Talks to an LLM?
A phenomenology of thinking with a model.
LLMs believe false statements even after explicit warnings that they're false
Fine-tuning tests show "bias... toward confidently representing the claims as true."
LLM is a learned distribution p on sequences of tokens. If you just sample a bunch of text from p willy nilly and put it in the training data and train you just get back p. But if you sample a bunch of text from p and throw out whatever's bad and train on what's left then you learn p(x|x is not bad)
There's growing experimental evidence¹˒²˒³˒⁴ that LLMs are unusually good at talking people out of entrenched false beliefs and conspiracy theories. ¹ pubmed.ncbi.nlm.nih.gov/39264999/ ² pubmed.ncbi.nlm.nih.gov/41179846/ ³ sciety.org/articles/activity/10.21203/rs… ⁴ connecticut.adl.org/resources/report/short-dialog…
Der Fisher
Unpopular opinion: LLMs are generally deradicalizing people whose beliefs are based on pure propaganda. I've been watching this happen slowly with some people I care about. ChatGPT is one of the reasons they finally accept that the NRA, OAN, and NewsMax are lying to them.
