Sublime
An inspiration engine for ideas
What does DeepSeek R1 & v3 mean for LLM data?
Contrary to some lazy takes I’ve seen, DeepSeek R1 was trained on a shit ton of human-generated data—in fact, the DeepSeek models are setting records for the disclosed amount of post-training data for open-source models:
- 600,000 reasoning data... See more
i helped grow google pay from 0 → 100M users in 3 yrs.
worked with @danariely + the world's top behavioral scientists.
here’s what actually worked (and what you should steal for your own consumer app):
1. variable rewards: humans crave unpredictable wins.
scratch... See more
Sherry Jiangx.com