Optimizing AI Agents: Prompts vs. Weights (GEPA, GRPO) | Stanford CS329Z Lecture 9 with Michael Ryan

Optimizing AI Agents: Prompts vs. Weights (GEPA, GRPO) | Stanford CS329Z Lecture 9 with Michael Ryan

Title history
  • Optimizing AI Agents: Prompts vs. Weights (GEPA, GRPO) | Stanford CS329Z Lecture 9 with Michael Ryan — observed on 20261117-2211 UTC (YouTube)
🎙 Michael Ryan 👥 1.4M 📅 November 17, 2026 ⏱ 80 min 👁 291 📄 expert opinion 🧭 2026-11-17
Available in: English (current) Français

Keywords

AI optimizationprompt optimizationweight optimizationGRPOGEPA

Summary

Michael Ryan, a Stanford CS PhD student, presents a lecture on optimizing AI agents by comparing prompt optimization and weight optimization. He begins by discussing the conditions required for optimization: data, system, and metrics. He then covers context optimization, including few-shot optimization with methods like bootstrapped few-shot and OPRO, which uses a language model to generate new prompts based on past scores. He also mentions reflective optimization methods like MIPROv2, TextGrad, and GEPA. The lecture then shifts to weight optimization, covering supervised fine-tuning on agent rollouts, and reinforcement learning algorithms such as PPO and GRPO, along with on-policy distillation. He discusses using agents as optimizers, like the Darwin Gödel Machine and autoregressive research. Finally, he provides practical guidance on when to choose prompt optimization versus weight tuning. Throughout, he emphasizes the importance of evals and data-driven optimization for last-mile adaptation.

140 words

Critical Evaluation

Value of the Information & Strength of the Argument

The lecture provides a comprehensive overview of optimization techniques for AI systems, with clear explanations and a structured framework. The arguments are well-supported by references to existing research and practical examples. The speaker’s expertise is evident, and he effectively illustrates the trade-offs between context and weight optimization. The presentation is logically organized, progressing from basic concepts to advanced methods, and includes interactive polls and Q&A to engage the audience. The depth of coverage is appropriate for a graduate-level course, making it valuable for both students and practitioners.

Scientific Rigor, Source Quality, Title Accuracy

The lecture references several key papers and methods, including OPRO, GRPO, and GEPA, with proper attribution. The content is based on established research and the speaker’s own contributions to DSPy and prompt optimization. The title accurately reflects the content, and the lecture is part of a Stanford course, which adds to its credibility. The use of examples and demonstrations enhances the pedagogical value. Overall, the sources are reliable, and the presentation adheres to academic standards.

177 words

Title / Content Match

The title precisely describes the lecture's focus on comparing prompt and weight optimization, mentioning specific algorithms (GEPA, GRPO) and the Stanford course context.

Quality & Reliability

9/10

The lecture is delivered by a Stanford CS PhD student and researcher actively contributing to prompt optimization, referencing established papers and methods, and is part of a graduate course at Stanford, ensuring a high level of expertise and reliability.

Key Moments

Cited Sources

Contribution & Novelties

This lecture offers an original synthesis of optimization strategies for AI agents, bridging context optimization (prompts) and weight optimization (fine-tuning). It introduces several modern algorithms and their trade-offs, underscored by the speaker’s own research in prompt optimization. The practical guidance based on evaluation setups is particularly valuable for practitioners deploying AI systems.

Pour aller plus loin :

138 words

Radar Profile

The radar profile shows uniformly high scores across all dimensions, reflecting a balanced and comprehensive lecture. Information quantity and quality are top-notch, technical depth is high, and reliability is assured by the Stanford affiliation and expert speaker.

Reliability 9/10