Truth Checker: Generative Large Language Models and Hallucinations with Chandra Khatri

Truth Checker: Generative Large Language Models and Hallucinations by Chandra Khatri navigates the era of artificial intelligence, understanding the capabilities and limitations of these technologies becomes crucial, especially their tendency to produce confident but inaccurate information, known as hallucinations. This session aims to unveil the mechanisms of Truth Checker models that pinpoint and mitigate these hallucinations, paving the way for more reliable applications of LLMs in both enterprise and end-user contexts without the need for human oversight.

#MachineLearning #DeepLearning #AI #ArtificialIntelligence #NLP #DataScience #DataVisualization #GitHub #DataEngineering #NaturalLanguageProcessing #ReinforcementLearning #TechTalks

Timecodes:
0:00 - Intro
1:15 - Why is Language Understanding hard for Machines?
3:58 - Even for humans language undertaking is one of the hardest tasks
5:50 - But ChatGPT can do these tasks. How is it trained?
8:27 - What’s Happened in the last - 5 Months
10:45 - Hallucinations, When LLS make stuff up
13:50 - If not ChatGPT, which LLMs can I use?
20:37 - Running LLMs Internally can be costly
25:04 - What about Halluncations?
27:08 - ELMAR Response Generation and Truth Checker Performance
34:00 - Q&A

You can also follow ODSC on: