method · first seen 10 Oct 2026
weight-only quantization
Reducing model size by quantizing only the weights of a neural network to low bit widths while keeping activations in higher precision.
Mentions per week
PapersCodeDiscussion
Earliest sources
- SAGE: Pushing 2-Bit Vector-Quantized LLMs toward Real-World UsabilityPapers · 10 Oct 2026 · Xianglong Yan, Zenan Wu, Kewei Zhang, Kaicheng Yang, Yulun Zhang
- FIGFLEX: A Mixed-Precision FP-INT GEMM Accelerator Co-Designed with Low-Bit LLM QuantizationPapers · 21 Oct 2026 · Faraz Tahmasebi, Haocheng Xu, Andrew Trauger, Yomal Perera, Zhenyu Tang, Nael Ch…
Latest sources
- FIGFLEX: A Mixed-Precision FP-INT GEMM Accelerator Co-Designed with Low-Bit LLM QuantizationPapers · 21 Oct 2026 · Faraz Tahmasebi, Haocheng Xu, Andrew Trauger, Yomal Perera, Zhenyu Tang, Nael Ch…
- SAGE: Pushing 2-Bit Vector-Quantized LLMs toward Real-World UsabilityPapers · 10 Oct 2026 · Xianglong Yan, Zenan Wu, Kewei Zhang, Kaicheng Yang, Yulun Zhang