REVIEW 33 cited by
A Survey of Quantization Methods for Efficient Neural Network Inference
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
A Survey of Quantization Methods for Efficient Neural Network Inference
read the original abstract
As soon as abstract mathematical computations were adapted to computation on digital computers, the problem of efficient representation, manipulation, and communication of the numerical values in those computations arose. Strongly related to the problem of numerical representation is the problem of quantization: in what manner should a set of continuous real-valued numbers be distributed over a fixed discrete set of numbers to minimize the number of bits required and also to maximize the accuracy of the attendant computations? This perennial problem of quantization is particularly relevant whenever memory and/or computational resources are severely restricted, and it has come to the forefront in recent years due to the remarkable performance of Neural Network models in computer vision, natural language processing, and related areas. Moving from floating-point representations to low-precision fixed integer values represented in four bits or less holds the potential to reduce the memory footprint and latency by a factor of 16x; and, in fact, reductions of 4x to 8x are often realized in practice in these applications. Thus, it is not surprising that quantization has emerged recently as an important and very active sub-area of research in the efficient implementation of computations associated with Neural Networks. In this article, we survey approaches to the problem of quantizing the numerical values in deep Neural Network computations, covering the advantages/disadvantages of current methods. With this survey and its organization, we hope to have presented a useful snapshot of the current research in quantization for Neural Networks and to have given an intelligent organization to ease the evaluation of future research in this area.
Forward citations
Cited by 33 Pith papers
-
The Geometry of LLM Quantization: GPTQ as Babai's Nearest Plane Algorithm
GPTQ is equivalent to Babai's nearest plane algorithm for CVP on the Hessian lattice of layer inputs, yielding geometric interpretation, inherited error bounds, and improved clipping-free quantization with GPU kernels.
-
The Complexity of Verifying Feedforward Neural Networks in Quantised Settings
Verification of fixed-precision quantized FNNs is NP-complete under both LP and BV specifications, matching the rational case, while dynamic quantization with BV specs has established upper bounds complementing known ...
-
When Bits Break Recourse: Counterfactual-Faithful Quantization
CFQ trains quantizer parameters and mixed-precision allocation to preserve counterfactual recourse validity, cost, and direction on Adult, German Credit, and COMPAS while matching accuracy of standard quantizers.
-
Characterizing Learning in Deep Neural Networks using Tractable Algorithmic Complexity Analysis
QuBD extends algorithmic complexity estimation to quantized DNN weights, revealing that complexity decreases during learning, increases with overfitting, follows grokking patterns, and correlates with generalization.
-
DPQuant: Efficient and Differentially-Private Model Training via Dynamic Quantization Scheduling
DPQuant uses epoch-wise probabilistic layer rotation and DP loss sensitivity to quantize only a changing subset of layers, reducing accuracy degradation from quantization noise in DP-SGD and delivering up to 2.21x thr...
-
GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers
GPTQ quantizes 175B-parameter GPT models to 3-4 bits per weight in one shot using approximate second-order information, achieving negligible accuracy degradation and 3-4x inference speedups.
-
LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale
LLM.int8() performs 8-bit inference for transformers up to 175B parameters with no accuracy loss by combining vector-wise quantization for most features with 16-bit mixed-precision handling of systematic outlier dimensions.
-
Lyapunov-Guided Training for Hardware-Safe Neural Networks Under Fixed-Point Arithmetic
Monotone Lyapunov projection of layerwise hidden-state energy suppresses two's-complement overflow under wrapping fixed-point QAT/PTQ, recovering 86.55% MNIST accuracy where unconstrained models collapse to chance.
-
SharQ: Bridging Activation Sparsity and FP4 Quantization for LLM Inference
SharQ combines input-adaptive N:M sparsity and FP4 quantization via sparse backbone plus dense residual, recovering 43-63% of the NVFP4-to-FP16 accuracy gap on Llama and Qwen models without calibration or retraining.
-
Quantizing Time-Series Models As Dynamical Systems: Trajectory-Based Quantization Sensitivity Score
Introduces TQS metric and TQS-PTQ framework that uses dynamical-systems stability to enable a priori, calibration-free mixed-precision post-training quantization for time-series models.
-
Minimum Distortion Quantization with Specified Output Distribution
Derives optimal quantizer form X=σ(F^{-1}(F_W(W))) with permutation σ minimizing MMSE under specified output distribution P_X, using majorization.
-
Machine learning enables experimental access to photon-by-photon arrival times in scintillation detectors
Deep learning extracts photon-by-photon arrival times from scintillation detector waveforms using unsupervised training with a physically informed model, enabling improved timing resolution and photon classification i...
-
Prompt2Fingerprint: Plug-and-Play LLM Fingerprinting via Text-to-Weight Generation
P2F generates low-rank parameter increments for LLM fingerprinting directly from textual descriptions in a single forward pass.
-
Rethink the Role of Neural Decoders in Quantum Error Correction
Neural decoders for surface-code QEC achieve practical microsecond FPGA latency when trained on large datasets with appropriate inductive biases and INT4 quantization, rather than relying on architectural complexity.
-
LoKA: Low-precision Kernel Applications for Recommendation Models At Scale
LoKA enables practical FP8 use in numerically sensitive large recommendation models via online profiling of activations, reusable model modifications for stability, and dynamic kernel dispatching.
-
LoKA: Low-precision Kernel Applications for Recommendation Models At Scale
LoKA enables practical FP8 use in numerically sensitive large recommendation models via profiling, model adaptations, and runtime kernel orchestration.
-
Litespark Inference For CPUs: Ultra-Fast SIMD Framework for Ternary (1.58-bit) Language Models
Litespark-Inference delivers custom SIMD kernels for ternary LLMs achieving up to 95.81x throughput versus PyTorch on CPUs by using integer addition/subtraction instead of floating-point math.
-
Litespark Inference For CPUs: Ultra-Fast SIMD Framework for Ternary (1.58-bit) Language Models
Custom SIMD kernels for ternary LLMs deliver 9.2x faster time-to-first-token, 52x higher throughput, and 14x memory reduction versus standard PyTorch on Apple Silicon and similar CPUs.
-
Initialisation Determines the Basin: Efficient Codebook Optimisation for Extreme LLM Quantization
Output-aware EM initialization for codebooks in additive quantization avoids poor optimization basins and yields better 2-bit compressed LLMs across Llama and Qwen models.
-
You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations
TAQ estimates per-layer importance from hidden representations and output sensitivity on task calibration data to allocate mixed precision in a training-free PTQ setting, outperforming task-agnostic baselines on accur...
-
AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration
AWQ quantizes LLM weights to low bits by scaling salient channels based on activation statistics, outperforming prior methods on language, coding, math, and multi-modal benchmarks.
-
Quantization in Federated Learning: Methods, Challenges and Future Directions
This survey introduces a taxonomy for quantization in federated learning organized around client heterogeneity, aggregation consistency, non-IID robustness, privacy integration, and hardware co-optimization, while ana...
-
PALUTE: Processing-In-Memory Acceleration via Lookup Table for Edge LLM Inference
PALUTE is a new PIM accelerator using in-DRAM LUTs on M3D DRAM that reports 1264 TPS at 0.16 W with 12.8x energy efficiency gains over CHIME for quantized edge LLM inference.
-
The Thermodynamic Costs of Simple Linear Regression
Thermodynamic lower bounds are approximated for exact and SGD linear regression, producing energy-aware scaling laws for optimal training dataset size given a target generalization error.
-
LoKA: Low-precision Kernel Applications for Recommendation Models At Scale
LoKA claims to make FP8 practical for large recommendation models via statistical probing, model adaptations, and accuracy-aware kernel dispatch.
-
Sustainability Is Not Linear: Quantifying Performance, Energy, and Privacy Trade-offs in On-Device Intelligence
Empirical case study on a flagship Android device profiles energy, latency, and quality trade-offs across eight LLMs, revealing a quantization energy paradox and identifying mid-sized models as practical sweet spots.
-
PicoSAM3: Real-Time In-Sensor Region-of-Interest Segmentation
A 1.3M-parameter CNN with ROI-implicit prompting and SAM3 distillation reaches ~65% mIoU on COCO/LVIS and 11.82 ms INT8 inference fully in-sensor on the Sony IMX500.
-
MixFrag: Fragility-Guided Mixed-Precision Post-Training Quantization for Vision Transformers
KL-isolation fragility plus MCKP bit allocation yields mixed-precision ViT PTQ that lags recent ImageNet PTQ but reports large COCO AP gains at MP3/MP3.
-
$\text{Log}_\text{b}$Quant: Quantizing Language Models in Logarithmic Space
Log_b Quant is an adjustable-base logarithmic quantization technique that outperforms tensor-wise asymmetric linear quantization at 4-bit precision on language model benchmarks while providing memory savings.
-
On the Expressive Power of Weight Quantization in Large Language Models
Weight-quantized LLMs retain universal approximation up to 1.58 bits with expressive collapse below it and polynomial degradation in capacity as bit count falls.
-
Memristor Technologies for Dynamic Vision Sensors: A Critical Assessment and Research Roadmap
A structured review concludes that end-to-end DVS-memristor integration for analog in-memory event-driven computing remains an open challenge at TRL 2-5, with half of surveyed applications resting on projections rathe...
-
Harnessing Photonics for Machine Intelligence
Photonic computing can reshape AI acceleration through optical bandwidth and parallelism, but requires cross-layer co-design and electronic-photonic design automation to move from prototypes to scalable systems.
-
K-Quantization and its Impact on Output Performance
Empirical evaluation of quantization effects on eight LLMs across bit widths, showing performance generally declines at lower precision but with model-size-dependent resilience and acceptable accuracy at 2 bits for ma...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.