Cover Image

GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Dadgarnia, Alireza, Tabesh, Soroush, Nikdan, Mahdi, Helcig, Michael, Kurtic, Eldar, Kleinegger, Maximilian, Alistarh, Dan
Format:	Preprint
Published:	2026
Subjects:	Computation and Language Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Statistically-Lossless Quantization of Large Language Models
by: Helcig, Michael, et al.
Published: (2026)

RoSA: Accurate Parameter-Efficient Fine-Tuning via Robust Adaptation
by: Nikdan, Mahdi, et al.
Published: (2024)

HALO: Hadamard-Assisted Lower-Precision Optimization for LLMs
by: Ashkboos, Saleh, et al.
Published: (2025)

MatGPTQ: Accurate and Efficient Post-Training Matryoshka Quantization
by: Kleinegger, Maximilian, et al.
Published: (2026)

QuEST: Stable Training of LLMs with 1-Bit Weights and Activations
by: Panferov, Andrei, et al.
Published: (2025)

ECO: Quantized Training without Full-Precision Master Weights
by: Nikdan, Mahdi, et al.
Published: (2026)

EvoPress: Accurate Dynamic Model Compression via Evolutionary Search
by: Sieberling, Oliver, et al.
Published: (2024)

CAGE: Curvature-Aware Gradient Estimation For Accurate Quantization-Aware Training
by: Tabesh, Soroush, et al.
Published: (2025)

Mathador-LM: A Dynamic Benchmark for Mathematical Reasoning on Large Language Models
by: Kurtic, Eldar, et al.
Published: (2024)

Model Compression with Exact Budget Constraints via Riemannian Manifolds
by: Helcig, Michael, et al.
Published: (2026)

Quartet II: Accurate LLM Pre-Training in NVFP4 by Improved Unbiased Gradient Estimation
by: Panferov, Andrei, et al.
Published: (2026)

Error Feedback Can Accurately Compress Preconditioners
by: Modoranu, Ionut-Vlad, et al.
Published: (2023)

"Give Me BF16 or Give Me Death"? Accuracy-Performance Trade-Offs in LLM Quantization
by: Kurtic, Eldar, et al.
Published: (2024)

Panza: Design and Analysis of a Fully-Local Personalized Text Writing Assistant
by: Nicolicioiu, Armand, et al.
Published: (2024)

Quartet: Native FP4 Training Can Be Optimal for Large Language Models
by: Castro, Roberto L., et al.
Published: (2025)

MicroAdam: Accurate Adaptive Optimization with Low Space Overhead and Provable Convergence
by: Modoranu, Ionut-Vlad, et al.
Published: (2024)

Efficient Data Selection at Scale via Influence Distillation
by: Nikdan, Mahdi, et al.
Published: (2025)

DarwinLM: Evolutionary Structured Pruning of Large Language Models
by: Tang, Shengkun, et al.
Published: (2025)

Gumbel-MPNN: Graph Rewiring with Gumbel-Softmax
by: Hoffmann, Marcel, et al.
Published: (2025)

Bridging the Gap Between Promise and Performance for Microscaling FP4 Quantization
by: Egiazarian, Vage, et al.
Published: (2025)

GSQ-Tuning: Group-Shared Exponents Integer in Fully Quantized Training for LLMs On-Device Fine-tuning
by: Zhou, Sifan, et al.
Published: (2025)

Universal Adversarial Suffixes Using Calibrated Gumbel-Softmax Relaxation
by: Soor, Sampriti, et al.
Published: (2025)

LLMQ: Efficient Lower-Precision Pretraining for Consumer GPUs
by: Schultheis, Erik, et al.
Published: (2025)

Accurate Compression of Text-to-Image Diffusion Models via Vector Quantization
by: Egiazarian, Vage, et al.
Published: (2024)

MatryoshkaLoRA: Learning Accurate Hierarchical Low-Rank Representations for LLM Fine-Tuning
by: Modoranu, Ionut-Vlad, et al.
Published: (2026)

LRW-Persian: Lip-reading in the Wild Dataset for Persian Language
by: Taghizadeh, Zahra, et al.
Published: (2025)

Apertus LLM Family Expansion via Distillation and Quantization
by: Panferov, Andrei, et al.
Published: (2026)

Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment
by: Agarwalla, Abhinav, et al.
Published: (2024)

Behemoth: Benchmarking Unlearning in LLMs Using Fully Synthetic Data
by: Iofinova, Eugenia, et al.
Published: (2026)

Federated Variational Preference Alignment with Gumbel-Softmax Prior for Personalized User Preferences
by: Koo, Jabin, et al.
Published: (2026)

Joint Beamforming and Integer User Association using a GNN with Gumbel-Softmax Reparameterizations
by: Lyu, Qing, et al.
Published: (2025)

Gumbel-Softmax Flow Matching with Straight-Through Guidance for Controllable Biological Sequence Generation
by: Tang, Sophia, et al.
Published: (2025)

Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?
by: Zverev, Egor, et al.
Published: (2024)

Optimal sensor placement for the reconstruction of ocean states using differentiable Gumbel-Softmax sampling operator
by: Chapron, Oscar, et al.
Published: (2026)

Conditional Gumbel-Softmax for constrained feature selection with application to node selection in wireless sensor networks
by: Strypsteen, Thomas, et al.
Published: (2024)

SimA: Simple Softmax-free Attention for Vision Transformers
by: Koohpayegani, Soroush Abbasi, et al.
Published: (2022)

Speculative Decoding Speed-of-Light: Optimal Lower Bounds via Branching Random Walks
by: Pankratov, Sergey, et al.
Published: (2025)

Extreme Compression of Large Language Models via Additive Quantization
by: Egiazarian, Vage, et al.
Published: (2024)

Optimal design of frame structures with mixed categorical and continuous design variables using the Gumbel-Softmax method
by: Ebrahimi, Mehran, et al.
Published: (2024)

Pushing the Limits of Large Language Model Quantization via the Linearity Theorem
by: Malinovskii, Vladimir, et al.
Published: (2024)