LangDiversity: Software to Identify LLM Errors - Diversity Measures and Prompt Selection
Offered By: Neuro Symbolic via YouTube
Course Description
Overview
Explore a software tool designed to identify errors in large language model outputs through an 11-minute video tutorial. Learn about LangDiversity, an implementation of domain-independent "diversity measures" used to assess uncertainty in language model results. Discover the concept of diversity measures, their utility implementation, and see code examples. Delve into prompt selection techniques, including few-shot use cases, and their implementation. Access the LangDiversity package via pip, explore its GitHub repository, and read the associated research paper. Gain insights into addressing challenges like hallucination in language models and enhancing output reliability.
Syllabus
Intro:
Diversity Measure Concept:
Diversity Measure Utility Implementation:
Diversity Measure Utility Code Example:
Prompt Selection Concept:
Prompt Selection Use Case Few-shot:
Prompt Selection Utility Implementation:
Prompt Selection Utility Code Example:
Outro & References:
Taught by
Neuro Symbolic
Related Courses
Visual Concept Grounding for Lifelong LearningNeuro Symbolic via YouTube Human-Aware Metacognitive AI - Lecture on Planning with Incomplete Model Knowledge
Neuro Symbolic via YouTube Deep Ontological Networks: An Overview and Implementation
Neuro Symbolic via YouTube Scene Graphs: Overview and Current Research
Neuro Symbolic via YouTube Language Agents with LLMs - Evolutionary Step in Artificial Intelligence
Neuro Symbolic via YouTube