tweetindex

Diyi Yang @Diyi_Yang · 02 Sep 2026

The lectures will cover all the pieces to build agents from training data, RAG, agent design patterns/harnesses, evals, etc. Hw1: students will build an agentic harness from scratch. Hw2: students will create novel evals that challenge frontier models. Finally we’ll have
16 107Views
84Likes
3Reposts
4Replies
0Quotes
38Bookmarks

Is that a lot?

6.62×vs this author's median2 432 views is typical
57Percentile for this authorof 7 recent posts
9.84×vs 10K–100K median1 637 views is typical
63.61%Reachviews ÷ followers
0.56%Engagement rateof viewers reacted

Compare with the benchmark table →

Open on X →