Stratified Sparse Attention narrows the sparse-to-dense gap
On Ministral-3-8B at 50x sparsity, SSA lifts RULER-HARD from 63.8% to 69.3% at 128K.
50
Product
FlashInfer
- Mentions · 7d1
- Mentions · 30d2
- All mentions2
- First seenOct 3, 2026
- Last mentioned6d ago
Timeline
2 articles · newest firstPrime Intellect Launches Prime Inference: Serverless and Reserved Serving for Frontier Open Models
Prime Intellect launched Prime Inference, serverless and reserved serving for open models such as GLM-5.3.
52
Appears with
- NVIDIA Dynamoproduct · 1
- vLLMproduct · 1
- GLM-5.3ai model · 1
- Ministral-3-8Bai model · 1
- Prime Intellectorganization · 1
- Mooncakeproduct · 1
- GB200 NVL72product · 1
- NVIDIAvendor · 1
- SemiAnalysisorganization · 1
- Prime Inferenceproduct · 1
- OpenRouterorganization · 1
- Inferactorganization · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.