The leading labs seem to actually care about humanity

[Schefter] NFC Championship Game: Washington Commanders at Philadelphia Eagles, next Sunday, Jan. 26, 3 p.m. ET, FOX.

Second Half Game Thread: Los Angeles Rams (10-7) at Philadelphia Eagles (14-3)

Trump was the one who in '20 tried to ban tiktok. Can anyone explain what is happening?

Game Thread: Los Angeles Rams (10-7) at Philadelphia Eagles (14-3)

Search trend for VPN skyrockets as Tiktok shuts down in US

[Highlight] WAS vs DET - Detroit called for 12 men on the field during Commanders 4th down attempt

Second Half Game Thread: Washington Commanders (12-5) at Detroit Lions (15-2)

Google just released a new architecture

[D] Titans: a new seminal architectural development?

Gwern on OpenAIs O3, O4, O5

People outside of this subreddit are still in extreme denial. World is cooked rn

Mark Zuckerberg believes in 2025, Meta will probably have a mid-level engineer AI that can write code, and over time it will replace people engineers.

[Wildcard] Game Thread: Green Bay Packers @ Philadelphia Eagles

AI Explained's Simple Bench has been updated with o1-12-17 (high) at 40.1%%, still lower than o1-preview's 41.7%

[R] rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking

Microsoft says with rStar-Math, it has demonstrated that small language models (SLMs) can rival or even surpass the math reasoning capability of OpenAI o1.

Train a 7B model that outperforms GPT-4o ?

Roon (OpenAI) and Logan (Google) have a disagreement

LiveBench-2024-06-24 (Initial version) vs 2024 end of year (6 months, 8 days later). SOTA Global score increase of 14.51. SOTA biggest increase was reasoning: 27.58. Caveat: Reasoning models including o1 are still poor at spatial and compositional reasoning. Livebench should add these two types.

What do you think about explainable ai?

Will expanding the H1B program get us closer to AGI?

Thoughts on the eve of AGI

Thoughts on OpenAI's o3

How do you define rewards for RL on chain of thought reasoning? Trying to understanding a bit more into how o3 from OpenAI was trained.