Quantization-Aware Pre-Training with Constrained Empirical Weight Distribution

arXiv:2609.32659 · cs.LG · Submitted 2026-09-26 · Read on arXiv

cs.LG

Submitted: 2026-09-26

Updated: 2026-09-26

Code: https://github.com/1733116199/cewt

Terminology

• Quantization — 64× in this paper · explained in 12 episodes
• Adam — 43× in this paper · explained in 2 episodes
• Accuracy — 18× in this paper · explained in 1 episode
• Perplexity — 18× in this paper · explained in 3 episodes
• Periodic Interpolation — 16× in this paper
• Dampening Loss — 14× in this paper
• Hadamard Transform — 8× in this paper
• Rounding Boundary Mass — 8× in this paper
• Ablation — 7× in this paper · explained in 1 episode
• Ablation Studies — 7× in this paper · explained in 1 episode
• Quantization-Aware Pre-Training — 6× in this paper
• Hypersphere Optimization — 5× in this paper
• Distillation — 3× in this paper · explained in 2 episodes
• Iterative Freezing — 3× in this paper
• Pseudo-Quantization Noise — 3× in this paper
• Quantization-Aware Training — 3× in this paper · explained in 3 episodes
• Regularization — 3× in this paper · explained in 3 episodes
• Straight-Through Estimator — 3× in this paper
• Transformer — 3× in this paper · explained in 2 episodes
• Conference on Computer Vision and Pattern Recognition — 2× in this paper
• Confidence-Guided Annealing — 2× in this paper
• Pruning — 2× in this paper · explained in 4 episodes
• Scaling Laws — 2× in this paper · explained in 4 episodes
• Core — 1× in this paper · explained in 2 episodes
• Data Augmentation — 1× in this paper · explained in 7 episodes
• Edge Devices — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Loss Function — 1× in this paper · explained in 2 episodes
• Mixture of Experts — 1× in this paper · explained in 2 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Self-Distillation — 1× in this paper · explained in 7 episodes
• Sparsity — 1× in this paper · explained in 4 episodes
• TENSOR-WISE QUANTIZATION GRID WITH CHANNEL-WISE SCALING — 1× in this paper
• While Post-Training Quantization — 1× in this paper

Sources

• Quartet: Native FP4 Training Can Be Optimal for Large Language Models
• Symbolic Discovery of Optimization Algorithms
• A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets
• Training Compute-Optimal Large Language Models
• Adam: A Method for Stochastic Optimization
• WinQ: Accelerating Quantization-Aware Training of Language Models Around Saddle Points
• The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
• Squat: Quant Small Language Models on the Edge
• LLaMA: Open and Efficient Foundation Language Models
• A Spectral Condition for Feature Learning

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks