mlsanityData to 7 Oct 2026

With Shared Microexponents, A Little Shifting Goes a Long Way

arXiv 2302.08007 ↗First posted 16 Feb 2023cs.LG cs.AI cs.ARcited by 47

Not on the map. Why?

Abstract (arXiv)

This paper introduces Block Data Representations (BDR), a framework for exploring and evaluating a wide spectrum of narrow-precision formats for deep learning. It enables comparison of popular quantization standards, and through BDR, new formats based on shared microexponents (MX) are identified, which outperform other state-of-the-art quantization approaches, including narrow-precision floating-point and block floating-point. MX utilizes multiple levels of quantization scaling with ultra-fine scaling factors based on shared microexponents in the hardware. The effectiveness of MX is demonstrated on real-world models including large-scale generative pretraining and inferencing, and production-scale recommendation systems.

Key facts. Cited by 47 papers in the map (data through 7 Oct 2026). Builds on 28 of its 53 references that are in the map. First posted on arXiv 16 Feb 2023.

Builds on

28 of 53 references.

References · 28

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  4. 2020Language Models are Few-Shot Learnerscited by 32,199
  5. 2020Scaling Laws for Neural Language Modelscited by 6,411
  6. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,773
  7. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  8. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  9. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,853
  10. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,736
  11. 2021Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsitycited by 3,014
  12. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,471
16 more
  1. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  2. 2017Mixed Precision Trainingcited by 1,135
  3. 2021A Survey of Quantization Methods for Efficient Neural Network Inferencecited by 798
  4. 2022Efficiently Scaling Transformer Inferencecited by 529
  5. 2022DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scalecited by 400
  6. 2019Q-BERT: Hessian Based Ultra Low Precision Quantization of BERTcited by 355
  7. 2019Deep Learning Recommendation Model for Personalization and Recommendation Systemscited by 327
  8. 2022The case for 4-bit precision: k-bit Inference Scaling Lawscited by 247
  9. 2019A Study of BFLOAT16 for Deep Learning Trainingcited by 227
  10. 2022FP8 Formats for Deep Learningcited by 187
  11. 2018Training Deep Neural Networks with 8-bit Floating Point Numberscited by 163
  12. 2021VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inferencecited by 59
  13. 2018Training DNNs with Hybrid Block Floating Pointcited by 39
  14. 2021FAST: DNN Training Under Variable Precision Block Floating Point with Stochastic Roundingcited by 31
  15. 2022DHEN: A Deep and Hierarchical Ensemble Network for Large-Scale Click-Through Rate Predictioncited by 12
  16. 2021Low-Precision Hardware Architectures Meet Recommendation Model Inference at Scalecited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 4; 2024: 18; 2025: 13; 2026: 12.

47 papers.

Sort:

By area:

  1. 2024LLM Inference Unveiled: Survey and Roofline Model Insightscited by 147
  2. 2023Microscaling Data Formats for Deep Learningcited by 132
  3. 2024GPTVQ: The Blessing of Dimensionality for LLM Quantizationcited by 78
  4. 2024LQER: Low-Rank Quantization Error Reconstruction for LLMscited by 39
  5. 2023FP8 versus INT8 for efficient deep learning inferencecited by 31
  6. 2024u-μP: The Unit-Scaled Maximal Update Parametrizationcited by 29
  7. 2024Tender: Accelerating Large Language Models via Tensor Decomposition and Runtime Requantizationcited by 22
  8. 2024MicroScopiQ: Accelerating Foundational Models through Outlier-Aware Microscaling Quantizationcited by 22
  9. 2023Scaling Down to Scale Up: A Cost-Benefit Analysis of Replacing OpenAI's LLM with Open Source SLMs in Productioncited by 21
  10. 2024QERA: an Analytical Framework for Quantization Error Reconstructioncited by 21
  11. 2024Learning from Students: Applying t-Distributions to Explore Accurate and Efficient Formats for LLMscited by 18
  12. 2025Block Rotation is All You Need for MXFP4 Quantizationcited by 18
35 more
  1. 2024Effective Interplay between Sparsity and Quantization: From Theory to Practicecited by 16
  2. 2025INT v.s. FP: A Comprehensive Study of Fine-Grained Low-bit Quantization Formatscited by 16
  3. 2024BitMoD: Bit-serial Mixture-of-Datatype LLM Accelerationcited by 12
  4. 2026LatentMoE: Toward Optimal Accuracy per FLOP and Parameter in Mixture of Expertscited by 12
  5. 2025Adaptive Prediction-Powered AutoEval with Reliability and Efficiency Guaranteescited by 9
  6. 2026HiFloat4 Format for Language Model Inferencecited by 9
  7. 2024DaCapo: Accelerating Continuous Learning in Autonomous Systems for Video Analyticscited by 8
  8. 2024Anda: Unlocking Efficient LLM Inference with a Variable-Length Grouped Activation Data Formatcited by 8
  9. 2025MX+: Pushing the Limits of Microscaling Formats for Efficient Large Language Model Servingcited by 8
  10. 2023ESPN: Memory-Efficient Multi-Vector Information Retrievalcited by 7
  11. 2025MixDiT: Accelerating Image Diffusion Transformer Inference with Mixed-Precision MX Quantizationcited by 7
  12. 2025RaZeR: Pushing the Limits of NVFP4 Quantization with Redundant Zero Remappingcited by 6
  13. 2025LO-BCQ: Block Clustered Quantization for 4-bit (W4A4) LLM Inferencecited by 6
  14. 2024Accelerating Communication in Deep Learning Recommendation Model Training with Dual-Level Adaptive Lossy Compressioncited by 4
  15. 2025Key, Value, Compress: A Systematic Exploration of KV Cache Compression Techniquescited by 4
  16. 2024SDQ: Sparse Decomposed Quantization for LLM Inferencecited by 3
  17. 2024BBS: Bi-directional Bit-level Sparsity for Deep Learning Accelerationcited by 3
  18. 2025P3-LLM: An Integrated NPU-PIM Accelerator for Edge LLM Inference Using Hybrid Numerical Formatscited by 3
  19. 2025Log-Normal Multiplicative Dynamics for Stable Low-Precision Training of Large Networkscited by 2
  20. 2025MOSS: Efficient and Accurate FP8 LLM Training with Microscaling and Automatic Scalingcited by 2
  21. 2024Scaling Laws For Mixed Quantizationcited by 1
  22. 2025Assigning Distinct Roles to Quantized and Low-Rank Matrices Toward Optimal Weight Decompositioncited by 1
  23. 2025Tensor-Compressed and Fully-Quantized Training of Neural PDE Solverscited by 1
  24. 20261-Bit Wonder: Improving QAT Performance in the Low-Bit Regime through K-Means Quantizationcited by 1
  25. 2026RPU -- A Reasoning Processing Unitcited by 1
  26. 2026UE5M3 FP4 Block Scaling for Stable Language Model Pretrainingcited by 1
  27. 2024SLaNC: Static LayerNorm Calibrationcited by 0
  28. 2024Error Diffusion: Post Training Quantization with Block-Scaled Number Formats for Neural Networkscited by 0
  29. 2026TriGen: NPU Architecture for End-to-End Acceleration of Large Language Models based on SW-HW Co-Designcited by 0
  30. 2026SemanticDialect: Semantic-Aware Mixed-Format Quantization for Video Diffusion Transformerscited by 0
  31. 2026The Thermodynamic Costs of Simple Linear Regressioncited by 0
  32. 2026GyRot: Leveraging Hidden Synergy between Rotation and Fine-grained Group Quantization for Low-bit LLM Inferencecited by 0
  33. 2026HBQ: Hierarchical Scaling Block Quantization with Hardware-Efficiency-Aware Design for Accurate LLM Inferencecited by 0
  34. 2026MicroQonv: Reshaping Convolution Tensors for Efficient Microscaling in Training and Inferencecited by 0
  35. 2026Format-Aware Fusion for Fast FP4 Pretrainingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026MiX: Micro-Inverted-Scaling for End-to-End Low-Bit Vision-Language Model Acceleration
  2. 2026BATQuant: Outlier-resilient MXFP4 Quantization via Learnable Block-wise Optimization
  3. 2026Stable FP4 Training via Transposition-Invariant Block Quantization
  4. 2025LittleBit: Ultra Low-Bit Quantization via Latent Factorization
  5. 2025MicroMix: Efficient Mixed-Precision Quantization with Microscaling Formats for Large Language Models
  6. 2024Accurate Block Quantization in LLMs with Outliers
  7. 2026MXSens: Sensitivity-Aware Mixed-Precision Quantization for Efficient LLM Inference
  8. 2026ScaleSweep: Accurate NVFP4 Post-Training Quantization of LLMs via Block Scale Initialization