Abstract
Trellis-coded quantization enables high-dimensional compression of large language model (LLM) weights at ultra-low bit widths without the exponentially large codebooks required by conventional vector quantization. Practical deployment, however, presents two challenges: reconstructing compressed weights at sufficient parallel throughput to avoid making dequantization an inference bottleneck, and maintaining quantization accuracy without costly incoherence transformations. We address these challenges with two complementary techniques. First, we introduce an ultra-low-complexity trellis dequantizer that uses a structured, hardware-efficient state-to-value mapping while preserving diverse reconstruction choices for trellis search. Second, we reformulate discrete trellis path optimization with a curvature-aware objective that reflects model sensitivity directly in the original coordinate space. Together, these techniques enable high-quality ultra-low-bit trellis quantization with inexpensive, highly parallel runtime reconstruction and without relying on Hadamard-based incoherence processing.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Que, X., Elkayam, N., Pyakurel, S., Pan, S., & Gope, D. (2026). XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization. https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization
MLA 9
Que, Xiaofan, et al. "XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization." https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization.
Chicago (author–date)
Que, Xiaofan, Nir Elkayam, Spandan Pyakurel, Shuokai Pan, and Dibakar Gope. 2026. "XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization." https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization.
Harvard
Que, X., Elkayam, N., Pyakurel, S., Pan, S. and Gope, D. (2026) 'XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization', Available at: https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization.
Vancouver
Que X, Elkayam N, Pyakurel S, Pan S, Gope D. XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization. https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization
IEEE
X. Que, N. Elkayam, S. Pyakurel, S. Pan, and D. Gope, "XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization," https://omanscience.com/en/articles/xor-trellis-ultra-low-complexity-dequantization-and-curvature-aware-hadamard-free-llm-quantization.