peachywrites·3d agoSnakes and TransformersA game of attention, memory and a few unexpected snakes.A response icon1A response icon1
Hemang Baldha·Sep 28Mamba vs Transformers: The Growing Notebook vs the Fixed State (and Mamba 3)Why does longer context still hurt latency and memory even when everyone says their model is “efficient”?
Navuluribalaji·Sep 26Forget About Transformers. Mamaba-New EraThe Quadratic Bottleneck: Standard Transformers scale at $O(N²)$ complexity. As sequence lengths increase, the Key-Value (KV) cache…
Muhammad Awais·Sep 25Beyond the Transformer: Why Mamba and the SSM Revolution Just Redrew the AI MapThe Hook: The Quadratic Ceiling and the Mamba Moment
Meryem Çiftçi·Sep 19Mamba’yı Derinlemesine Anlamak: Durum, Seçicilik ve Unutmanın GücüGiriş: Bir Önceki Yazıdan DevamA response icon1A response icon1
InAccredianbyIshanmandalworks·Sep 18Beyond Transformers: How State Space Models (Mamba) Could Redefine AI ArchitectureThe model powering your favorite AI tool has a dirty secret — it reads everything, every time.
Prabin Tiwari·Sep 15Inside Mamba: How Input-Dependent SSMs Solve the Transformer BottleneckThe quadratic bottleneck of Transformer has finally met its match. An inside look at the selective SSM architecture.
Karthik·Jul 24Mamba-3: The Architecture Quietly Challenging TransformersThe Bill Every LLM Eventually Has to Pay
juiceb0xc0de·Jul 4TMax-9B: Inside the Mamba Hybrid Built on QwenWeight watchers:A new series where I put open-source models on the scale and tell you what they’re actually carrying. I build brain atlases…
InCodeXbyRichard Shu·Jun 26Transformer Killers: State Space Models (Mamba/Jamba) ExplainedIf you’ve been following AI news lately, you’ve likely heard a lot of buzz about Mamba and Jamba. People are calling them “Transformer…
