🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

I have been thinking about this question for a while and need your help(!): what do you think is the biggest breakthrough in scaling (law) after Chinchilla?

nanoChinchilla. Reproducing Chinchilla-Optimal Scaling Phenomenon: Colab, 1 Hour, 100 Lines, + Beautiful Theory tinyurl.com/2saj6bkj

NanoDO is really great. Almost all my scaling experiments are using it for the past year and I am probably the no. 1 user. Highly recommend to give it a shot.

It was a pleasure working on Gemma 2. The team is relatively small but very capable. Glad to see it get released. On the origin of techniques: 'like Grok', 'like Mistral', etc. is a weird way to describe them as they all originated at Google Brain/DeepMind and the way they ended

fantastic (like really fantastic) work led by Katie Everett !

“It’s the people, not the projects …”

Lgtm

If you miss the NYTimes needle, especially one that is statistically uniform (blog.alexalemi.com/a-degree-of-ce…), you can use this page: alexalemi.com/random/electio… I whipped together to reason about the correlations between the swing states tonight as results come in.

“勿以恶小而为之，勿以善小而不为。惟贤惟德，能服于人” ("Do not do evil just because it is small; do not ignore doing good just because it is small. Only by being virtuous and moral can one be respected by others.") learned this in middle school, never forgot, passed it to my kids

Curious to hear thoughts about levels of AGI in math. My take (unchanged since march 2023): L1: solve IMO L3: breakthroughs in 2 areas (e.g., analysis/PDEs, algebra/number theory, geometry/topology), each ~ one Annals/JAMS/Acta paper L5: solve a Millennium problem L6: ...

It’s extremely exciting to see the frontiers of science and practice merge, yet equally disheartening to see when they become frontier of geopolitics. What wisdoms do AI offer to human?

Very cool! Congrats and thanks for trusting our codebase!+ Peter J. Liu (creator of nanodo), Roman Novak ,Katie Everett (pre-)lead of the codebase.

Wow, didn’t think I’d see this conjecture solved in my lifetime terrytao.wordpress.com/2025/02/25/the…

Really impressive work! Changing the exponent is highly nontrivial.