Explore the concept of semi-formal programming enabled by language model agents in this conference talk from ACM SIGPLAN. Delve into how large language models (LLMs) can bridge the gap between fully formal and informal programming approaches. Discover how semi-formal programs incorporate both structured data and informal elements, allowing for greater flexibility in capturing user intent. Learn about the potential applications of language model agents (LMAs) in executing semantically meaningful semi-formal information, guiding users towards full formalization, and generating programs based on specifications. Gain insights from speakers Josh Pollock, Arvind Satyanarayan, and Daniel Jackson as they discuss this innovative approach to creating more malleable programming environments.
Overview
Syllabus
[LIVE23] Language Model Agents Enable Semi-Formal Programming
Taught by
ACM SIGPLAN
Related Courses
-
Prompt Engineering with GPT: Programming for Custom Content
-
Bridging Informal and Formal Mathematical Reasoning with Neural Language Models
-
Programming Languages for AI Programming Agents
-
Autoformalization with Large Language Models - IPAM at UCLA
-
Programming Language as Language
-
Neural Language Models and Few-Shot Learning for Systematic Requirements Processing in Model-Driven Software Engineering
