Top all-time · #12142
Tri Dao
Personal blog · tridao.me/blogaiprogramming
Tri Dao's personal blog at tridao.me covers ai and programming. Since 2024, 6 different Hacker News readers have linked it 6 times, most often “Mamba-2 – State Space Duality”. It is still writing: the last post was on 2026-06-15, and you can follow it by feed.
Too few sampled comments (2) to say how often HN recommends it.?
Why HN links it
HN readers link Tri Dao's Mamba-2 blog series, which explains the Mamba-2 state space model from the paper's author. One commenter notes the series is not aimed at the general public.
Summarised by Claude from a sample of 2 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Homepage of Tri Dao.
Most-linked posts?
- Mamba-2 – State Space Duality2 people · 2 mentions · first linked 2024-06-03
What Hacker News says?
For those who are text oriented: https://newsletter.maartengrootendorst.com/p/a-visual-guide-... The paper author has a blog series but I don't think it's for general public https://tridao.me/blog/2024/mamba2-part1-model/
Comment linking /blog/2024/mamba2-part1-model/
on “Codestral Mamba”Submitted Mamba-2 – State Space Duality as a story · 151 points
HN discussionDao and Gu show that if you simplify Mamba so its state-space layer uses a diagonal matrix A_t that is a scalar times the identity matrix, i.e., A_t = a_t I, the state-space transformation can be expressed as a form of causal linear attention[a] by compounding coefficients a_1 ... a_t at each time step t. The equivalence of the simplified state-space layer and causal linear attention constitute the duality the autho…
Comment linking /assets/img/2024-05-31-mamba-2/pile_8k_mamba2-1400.webp
on “Mamba-2 – State Space Duality”Submitted ReplaySSM: Cache SSM Inputs, Not State (Tri Dao) as a story · 1 points
HN discussionSubmitted Gram Newton-Schulz: A Fast, Hardware-Aware Newton-Schulz Algorithm for Muon as a story · 1 points
HN discussionSubmitted FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-Precision as a story · 1 points
HN discussion
