Jayson Lynch

Working Papers SAGE: Self-play Adversarial Games Enhance Large Language Model Reasoning Capabilities

March 15, 2026

• Neil Thompson | Research Scientist, MIT Sloan School of Management and CSAIL
• Jayson Lynch | Research Scientist, CSAIL
• Hans Gundlach | Research Assistant, CSAIL

Saraswathy Amjith

Michael X. Wang

 

This paper introduces SAGE (Self-play Adversarial Games for Enhancement), a framework for improving LLM reasoning capabilities through adversarial self-play without human-curated data. 

Research Briefs Why “Meek,” Low-Budget Models Could Soon Match High-Budget Performance

January 30, 2026

• Neil Thompson | Research Scientist, MIT Sloan School of Management and CSAIL
• Hans Gundlach | Research Assistant, CSAIL
• Jayson Lynch | Research Scientist, CSAIL

Working Papers How fast are algorithms reducing the demands on memory? A survey of progress in space complexity

November 15, 2025

• Jayson Lynch | Research Scientist, CSAIL
• Neil Thompson | Research Scientist, MIT Sloan School of Management and CSAIL

Hayden Rome

Jeffery Li

Chirag Falor

 

In this paper, scholars present the first broad survey of how algorithmic progress has improved memory usage (space complexity) by analyzing 118 of the most important algorithm problems in computer science, reviewing the 800+ algorithms used to solve them.