🔥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

🚨New Paper Alert As a game company, KRAFTON AI is actively exploring how to apply LLM agents to video games. We present Orak—a foundational video gaming benchmark for LLM agents! Includes Pokémon, StarCraft II, Slay the Spire, Darkest Dungeon, Ace Attorney, and more in🧵

Some of the most impactful work you can do in academia isn’t cool new algos or novel architectures. It’s data research. Data research isn’t just dumping tokens into a json. It requires a ton of rigorous experimentation, algorithmic thinking, and actually talking to your models.

I find it surprising when I meet LLM folks that don't talk to the models they're training/evaluating. That is an incredibly fun aspect of the research AND is super informative. You're no longer training a CIFAR10 classifier you'd never use. This thing can literally talk to you!

saying this as a person that loves algorithmic research btw

seed optimization => acc goes up => ...

Why is gemini mean and why does it love analogies?

Pretraining tokens offer the densest PRM imaginable.

Best way to get incredibly good peer review on your papers is to have them go viral on twitter. It is however advisable to be near certain they are correct before doing so.

Gemini 2.5 is extremely good at explaining data, and coming up with theoretical hypotheses that are sound. It's also great at "idea" debugging when forming speculations about phenomena. I'm actually impressed. Better than o3.

If you're ever in Athens, I highly recommend the Museum of Ancient Greek Technology! maps.app.goo.gl/sngwM98ueKkroo… Early humanoid robots to pour water+wine, demonstrations of heat+steam+water weight to "magically" open 500 kg temple doors, armor, ... (cont)

Claude Opus 6 may replace me, but not this

Today, I’m launching a deeply personal project. I’m betting $100M that we can help computer scientists create more upside impact for humanity. Built for and by researchers, including Jeff Dean & Joelle Pineau on the board, Laude Institute catalyzes research with real-world impact.

🚨New Paper Alert🚨 Could generative agents powered by LLMs transform social science by accurately simulating human social behaviors at scale? We tested this possibility with virtual humans facing disease threats in "Infected Smallville." 🧵

2500 years standing. They knew how to build.