Cool? Useful? How about staying up to date with the AI community here on X:
RESEARCH
-
DeepSeek Releases Massive Open-Source Model as OpenAI Charges $200/Month
By
–
Do you actually realize what's happening? Deepseek just dropped a 1.6 trillion parameter open-source model featuring a 1 million token context window. OpenAI is charging $200/month. but China is giving it away completely free. Let that sink in.
-
Verified Follow Friday AI and Tech Influencers List
By
–
Here are the verified Follow Friday lists for your reader! Every handle has been checked directly against X — I flagged several wrong ones in the process (noted below). ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ #FollowFriday — CUTTING
-
QuatRoPE: A New Method for 3D Spatial Encoding in LLMs
By
–
How can we teach AI to truly understand 3D space? Researchers from Peking University & SUSTech present QuatRoPE. It's a new method that efficiently encodes the 3D relationships between objects for Large Language Models, using a scalable linear approach instead of a messy
-
Stanford Research on Virtual Beings and AI Agents
By
–
Stanford has already done the human research on virtual beings.
-
Game Theory Decision Making: Red Button Versus Collective Coordination
By
–
If you press red you instantly get $5000 If more than 50% press blue they get $5000 each If less than 50% press blue they all get $0 Ofc game theory works differently with money than lives but what would you pick?
-
DeepSeek-V4 Research Paper Released Focusing on Million-Token Context Efficiency
By
–
DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence Paper: https://
huggingface.co/deepseek-ai/De
epSeek-V4-Pro/blob/main/DeepSeek_V4.pdf
… -
DeepSeek-V4 Technical Report Details New Architecture and Optimizer
By
–
DeepSeek-V4 tech report is here! The team replaced standard attention with a hybrid compressed system and used a new Muon optimizer to make training faster and more stable. They also introduced improved layer connections to help the model handle complex reasoning more