Effective Sequential Monte Carlo for Language Model Probabilistic Programs
Offered By: ACM SIGPLAN via YouTube
Course Description
Overview
Explore the potential of Sequential Monte Carlo (SMC) for efficient inference in language model probabilistic programs through this conference talk from ACM SIGPLAN's LAFI'24. Delve into the LLaMPPL library, which facilitates rapid exploration of SMC algorithms for language modeling tasks and automates efficient implementation, including auto-batching of large language model calls. Gain insights into key design challenges in SMC, focusing on intermediate target and proposal distribution design. Examine three example models that demonstrate superior performance compared to state-of-the-art language models and constrained generation techniques across various tasks. Learn from the perspectives of Alexander K. Lew, Tan Zhi-Xuan, Gabriel Grand, Jacob Andreas, and Vikash K. Mansinghka as they discuss the integration of probabilistic programming with large language models to encode complex distributions beyond traditional prompting methods.
Syllabus
[LAFI'24] Effective Sequential Monte Carlo for Language Model Probabilistic Programs
Taught by
ACM SIGPLAN
Related Courses
Discrete Inference and Learning in Artificial VisionÉcole Centrale Paris via Coursera Teaching Literacy Through Film
The British Film Institute via FutureLearn Linear Regression and Modeling
Duke University via Coursera Probability and Statistics
Stanford University via Stanford OpenEdx Statistical Reasoning
Stanford University via Stanford OpenEdx