Author name: Edward J. Yoon

ITQ3_S: High-Fidelity 3-bit LLM Inference via Interleaved Ternary Quantization with Rotation-Domain Smoothing

Edward J. Yoon / April 1, 2026

arXiv:2603.27914v2 Announce Type: replace
Abstract: We present ITQ3_S (Interleaved Ternary Quantization — Specialized), a novel 3-bit weight quantization format for LLMs integrating TurboQuant (TQ), a rotation-domain strategy based on the Fast Walsh-…

cs.AI, cs.DC, cs.LG

ITQ3_S: High-Fidelity 3-bit LLM Inference via Interleaved Ternary Quantization with Rotation-Domain Smoothing

Edward J. Yoon / March 31, 2026

arXiv:2603.27914v1 Announce Type: new
Abstract: We present \textbf{ITQ3\_S} (Interleaved Ternary Quantization — Specialized), a novel 3-bit weight quantization format for large language models (LLMs) that integrates \textbf{TurboQuant (TQ)}, a rotati…