Building a Reasoning Machine - Ph.D. thesis | Edward Hu

An introduction to my PhD thesis on scaling in the service of reasoning and model-based ML.

Building a Reasoning Machine - Ph.D. thesis | Edward Hu Building a Reasoning Machine - Ph.D. thesis Tuesday. April 07, 2026 reasoning GFlowNet thesis This is the introduction of my PhD thesis. Despite the immense progress of ML systems in interactive and generative tasks [ Brown et al., 2020 ; Ho et al., 2020 ], the best systems to date still struggle with tasks that require deliberate and sequential reasoning, such as solving complex math problems or creating a detailed and coherent plan of action [ Wei et al., 2022 ; Kojima et al., 2022 ]. Several features of human intelligence are not yet p

saved by

• Dhruv Sheth

related reading

• MAI-Thinking-1: Building a Hill-Climbing Machinemicrosoft.ai
• State of RL for reasoning LLMs | A. Weersaweers.de
• Scaling in the service of reasoning & model-based ML | Yoshua Bengioyoshuabengio.org
• the-illusion-of-thinking.pdfml-site.cdn-apple.com
• DeepSeek-R1arxiv.org
• [2412.06769] Training Large Language Models to Reason in a Continuous Latent Spacearxiv.org
• [2201.11903] Chain of Thought Prompting Elicits Reasoning in Large Language Modelsarxiv.org
• pdfopenreview.net
• Training Large Language Models to Reason in a Continuous Latent Spacearxiv.org
• Reasoning Models Reason Well, Until They Don'tarxiv.org
• Base Models Know How to Reason, Thinking Models Learn Whenarxiv.org
• dennyzhou.github.io/LLM-Reasoning-Stanford-CS-25.pdfdennyzhou.github.io