Hallucination Is Inevitable: An Innate Limitation of Large Language Models

#1

Post by PerryCox » Fri, Feb 09, 2024, 1:37 PM UTC

Hallucination Is Inevitable: An Innate Limitation of Large Language Models
arxiv.org

Re: Hallucination Is Inevitable: An Innate Limitation of Large Language Models

#2

Post by rapatel0 » Fri, Feb 09, 2024, 1:48 PM UTC

This is pretty obvious no? LLMs are basically a lossy compression of their dataset. Being lossy, they will necessarily have error (hallucinations). Furthermore, the underlying data is human approximation of truth. Therefore it will have error as a well. Shall we publish a paper on the linked list?

Re: Hallucination Is Inevitable: An Innate Limitation of Large Language Models

#3

Post by rhelz » Fri, Feb 09, 2024, 2:36 PM UTC

rapatel0 wrote: post #2

This is pretty obvious no? LLMs are basically a lossy compression of their dataset. Being lossy, they will necessarily have error (hallucinations). Furthermore, the underlying data is human approximation of truth. Therefore it will have error as a well. Shall we publish a paper on the linked list?

Hmmm....couldn't you argue it the other way as well? I mean, a hallucination isn't just an incomplete fact present in the input set--it is a claim which wasn't in the input set.

That's not what we'd expect from a lossy compressor; we would expect incomplete or missing answers, but we wouldn't expect it to give us claims which were not in the input set.