AI & machine learning2025 · shared on Hacker News
DeepSeek-V2: A Strong, Economical, and Efficient MOE Language Model
Explore DeepSeek-V2: A Strong, Economical, and Efficient MOE Language Model, its original source and Hacker News discussions. Discover community perspectives and save it to your reading list.
GO BEYOND THE ABSTRACT
The conversation around the idea.
Original Hacker News threads. Different perspectives, questions, and things worth thinking about.
Showing up to 100 related threads. A paper may appear in a comment rather than the original submission.
Links found in the discussion
Up to 50 recent source links, in HN item order.
YAlifatisk ·
Google Titans architecture, helping AI have long-term memory
587 points186 comments
Ythm ·
Xiaomi MiMo Reasoning Model
482 points193 comments
Yt55 ·
DeepSeek's multi-head latent attention and other KV cache tricks
292 points72 comments
Ytaiboku256 ·
Ask HN: Is anybody building an alternative transformer?
148 points93 comments
Ymooreds ·
Attention Wasn't All We Needed
130 points24 comments
Ymirrir ·
Zebra-Llama – Towards efficient hybrid models
113 points61 comments
Y0xkato ·
How LLMs Work
4 points0 comments
Ysonabinu ·
DeepSeek-V2: A Strong, Economical, and Efficient MOE Language Model
3 points0 comments