Stopping Hallucinations From Hurting Your LLMs - Part 2

Explore the critical issue of hallucinations in Large Language Models (LLMs) through this insightful 15-minute conference talk by Atindriyo Sanyal, founder and CTO of Galileo. Delve into the definition of hallucinations in modern LLM workflows and understand their impact on model outcomes and downstream consumers. Discover novel and efficient metrics and methods for early detection of hallucinations, aimed at preventing disinformation and poor or biased outcomes from LLMs. Learn how to increase trust in your LLM systems by addressing this crucial evaluation metric. Gain valuable insights from Sanyal's extensive experience in building large-scale ML platforms at companies like Uber and Apple.

Stopping Hallucinations From Hurting Your LLMs // Atindriyo Sanyal // LLMs in Prod Conference Part 2