mlsanityData to 14 Oct 2026

Block Rotation is All You Need for MXFP4 Quantization

arXiv 2511.04214 ↗First posted 6 Nov 2025cs.LG cs.CLcited by 20

Not on the map. Why?

Abstract (arXiv)

Large language models (LLMs) have achieved remarkable success, but their rapidly growing scale imposes prohibitive costs in memory, computation, and energy. Post-training quantization (PTQ) is a promising solution for efficient deployment, yet achieving accurate W4A4 quantization remains an open challenge. While most existing methods are designed for INT4 formats, the emergence of MXFP4 -- a new FP4 format with various hardware support (NVIDIA, AMD, Intel)-- raises questions about the applicability of current techniques. In this work, we establish a comprehensive benchmark of PTQ methods under the MXFP4 format. Through systematic evaluation, we find that methods like GPTQ consistently deliver strong performance, whereas rotation-based approaches, which are almost used by all state-of-the-art approaches, suffer from severe incompatibility with MXFP4. We further provide the first in-depth analysis of this conflict, tracing its root to a fundamental mismatch between MXFP4's PoT (power-of-two) block scaling and the redistribution of outlier energy via global rotation. Building on this insight, we propose a simple yet effective block rotation strategy that adapts rotation-based methods to MXFP4, leading to substantial accuracy improvements across diverse LLMs. Our findings not only offer clear guidance for practitioners but also set a foundation for advancing PTQ research under emerging low-precision formats.

Key facts. Cited by 20 papers in the map (data through 14 Oct 2026). Builds on 28 of its 34 references that are in the map. First posted on arXiv 6 Nov 2025.

Builds on

28 of 34 references.

References · 28

  1. 2019PIQA: Reasoning about Physical Commonsense in Natural Languagecited by 2,883
  2. 2018Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answeringcited by 2,164
  3. 2024Qwen2 Technical Reportcited by 2,120
  4. 2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,919
  5. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,386
  6. 2024QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMscited by 634
  7. 2024SpinQuant: LLM quantization with learned rotationscited by 471
  8. 2023OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Modelscited by 424
  9. 2023QuIP: 2-Bit Quantization of Large Language Models With Guaranteescited by 414
  10. 2024QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebookscited by 378
  11. 2024QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Servingcited by 208
  12. 2024DuQuant: Distributing Outliers via Dual Transformation Makes Stronger Quantized LLMscited by 167
16 more
  1. 2024FlatQuant: Flatness Matters for LLM Quantizationcited by 131
  2. 2023A Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directionscited by 105
  3. 2020Efficient Riemannian Optimization on the Stiefel Manifold via the Cayley Transformcited by 80
  4. 2024AffineQuant: Affine Transformation Quantization for Large Language Modelscited by 67
  5. 2024VPTQ: Extreme Low-bit Vector Post-Training Quantization for Large Language Modelscited by 57
  6. 2018A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Datasetcited by 55
  7. 2023With Shared Microexponents, A Little Shifting Goes a Long Waycited by 50
  8. 2024APTQ: Attention-aware Post-Training Mixed-Precision Quantization for Large Language Modelscited by 34
  9. 2023QuantEase: Optimization-based Quantization for Language Modelscited by 29
  10. 2024AMXFP4: Taming Activation Outliers with Asymmetric Microscaling Floating-Point for 4-bit LLM Inferencecited by 19
  11. 2025FPTQuant: Function-Preserving Transforms for LLM Quantizationcited by 16
  12. 2024BoA: Attention-aware Post-training Quantization without Backpropagationcited by 12
  13. 2025KurTail : Kurtosis-based LLM Quantizationcited by 12
  14. 2025LO-BCQ: Block Clustered Quantization for 4-bit (W4A4) LLM Inferencecited by 7
  15. 2025RSQ: Learning from Important Tokens Leads to Better Quantized LLMscited by 3
  16. 2025BBAL: A Bidirectional Block Floating Point-Based Quantisation Accelerator for Large Language Modelscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 19.

20 papers.

Sort:

By area:

  1. 2025WUSH: Near-Optimal Adaptive Transforms for LLM Quantizationcited by 15
  2. 2026ARCQuant: Boosting NVFP4 Quantization with Augmented Residual Channels for LLMscited by 15
  3. 2026Pushing the Limits of Block Rotations in Post-Training Quantizationcited by 11
  4. 2026Benchmarking Post-Training Quantization of Large Language Models under Microscaling Floating Point Formatscited by 10
  5. 2026DuQuant++: Fine-grained Rotation Enhances Microscaling FP4 Quantizationcited by 8
  6. 2026High-Rate Quantized Matrix Multiplication Icited by 7
  7. 2026BATQuant: Outlier-resilient MXFP4 Quantization via Learnable Block-wise Optimizationcited by 6
  8. 2026SOAR: Scale Optimization for Accurate Reconstruction in NVFP4 Quantizationcited by 4
  9. 2026TORQ: Two-Level Orthogonal Rotation for MXFP4 Quantizationcited by 3
  10. 2026Certain Head, Uncertain Tail: Expert-Sample for Test-Time Scaling in Fine-Grained MoEcited by 2
  11. 2026Unleashing Low-Bit Inference on Ascend NPUs: A Comprehensive Evaluation of HiFloat Formatscited by 2
  12. 2026LATMiX: Learnable Affine Transformations for Microscaling Quantization of LLMscited by 2
8 more
  1. 2026Quantizing Recursive Reasoning Modelscited by 2
  2. 2026FOCUS: FP4 Optimization via Coupled-Relaxation and Dual-Granularity Scalingcited by 2
  3. 2026ReSET: Accurate Latency-Critical NVFP4 Reasoning via Step-Aware Temperature Scalingcited by 0
  4. 2026SharQ: Bridging Activation Sparsity and FP4 Quantization for LLM Inferencecited by 0
  5. 2026Transforms for LLM Quantization: The Great Inversion and Format Co-Designcited by 0
  6. 2026PRQuant: Permutation Residual Quantization for Low-Overhead Inferencecited by 0
  7. 2026JustQuant: You Don't Need Smoothing, SVD, or Rotation for 4-Bit Activation Quantizationcited by 0
  8. 2026QuantForge: Discovering Residual Decompositions for MXFP4 Post-Training Quantizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Unveiling the Potential of Quantization with MXFP4: Strategies for Quantization Error Reduction
  2. 2025Bridging the Gap Between Promise and Performance for Microscaling FP4 Quantization
  3. 2025Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling
  4. 2025FP4 All the Way: Fully Quantized Training of LLMs
  5. 2026Pretraining large language models with MXFP4 on Native FP4 Hardware
  6. 2026Stable FP4 Training via Transposition-Invariant Block Quantization
  7. 2023ZeroQuant-FP: A Leap Forward in LLMs Post-Training W4A8 Quantization Using Floating-Point Formats
  8. 2026OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization