DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression
15 min1
Fathom: Per-Query Read Depth for Sparse Decoding over Offloaded KV Caches
1
SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking
2
DeepSeek AI Released DeepSeek-V4.1-Flash with 1M Context, FP4 KV Cache, and Cross-Layer Attention Reuse
4 min2
deepseek-ai/DeepSeek-V4.1-Flash — new model trending #28 on Hugging Face
10 min1
unsloth/Qwen3.8-Flash-Next-GGUF — new model trending #21 on Hugging Face
15 min1