-
From Score-and-Rank to System Evolution: The LLM-Agent Paradigm Shift in Recommender Systems
Why going all-in and replacing the whole recommendation stack with an LLM agent is usually the wrong move, and what should actually change across memory, reasoning, compute, reward, and simulation
-
Notes on correlation
Pearson's r, coefficient of determination, and how to interpret them
-
civic si 体验总结
半年人车体验：第十代思域Si的外观、内饰与驾驶感受
-
Food that I always want to try again!
a foodie's incomplete guide to restaurants worth a second visit
-
Lecture notes for Stat665
cross-entropy, softmax, and other highlights from Yale's Stat665
