aiqbench analysis
This research addresses a critical bottleneck in deploying Large Language Models on edge devices by enabling direct fine-tuning of quantized parameters without relying on continuous weights. By integrating error feedback and stateless seed replay, the proposed Quantized Evolution Strategies method significantly outperforms existing zeroth-order optimization techniques for discrete model spaces. This advancement paves the way for scaling high-precision LLMs entirely within low-memory constraints while maintaining adaptability to new tasks.
Quantized Evolution Strategies: High-precision Fine-tuning of Quantized LLMs at Low-precision Cost
arXiv:2602.03120v2 Announce Type: replace-cross Abstract: Post-Training Quantization (PTQ) is essential for deploying Large Language Models (LLMs) on memory-constrained devices, yet it renders models static and difficult to fine-tune. Standard fine-tuning paradigms, including Reinforcement Learning (RL), fundamentally rely on backpropagation and continuous weights to compute gradients. Thus they cannot be used on quantized models, where the parameter space is discrete and non-differentiable. While Evolution Strategies (ES) offer a backpropagation-free alternative, optimization of the quantized parameters can still fail due to vanishing or inaccurate gradient estimation. This paper introduces Quantized Evolution Strategies (QES), an optimization paradigm that performs full-parameter fine-tuning directly in the quantized space. QES is based on two innovations: (1) it integrates accumulated error feedback to preserve high-precision weight updating signals, and (2) it utilizes a stateless seed replay to reduce memory usage to low-precision inference levels. QES significantly outperforms the state-of-the-art zeroth-order fine-tuning methods on a variety of tasks, making direct fine-tuning for quantized models possible. It therefore opens up the possibility for scaling up LLMs entirely in the quantized space. The source code is available at https://github.com/dibbla/Quantized-Evolution-Strategies .
Abstract:Post-Training Quantization (PTQ) is essential for deploying Large Language Models (LLMs) on memory-constrained devices, yet it renders models static and difficult to fine-tune. Standard fine-tuning paradigms, including Reinforcement Learning (RL), fundamentally rely on backpropagation and continuous weights to compute gradients. Thus they cannot be used on quantized models, where the parameter space is discrete and non-differentiable. While Evolution Strategies (ES) offer a backpropagation-free alternative, optimization of the quantized parameters can still fail due to vanishing or inaccurate gradient estimation. This paper introduces Quantized Evolution Strategies (QES), an optimization paradigm that performs full-parameter fine-tuning directly in the quantized space. QES is based on two innovations: (1) it integrates accumulated error feedback to preserve high-precision weight updating signals, and (2) it utilizes a stateless seed replay to reduce memory usage to low-precision inference levels. QES significantly outperforms the state-of-the-art zeroth-order fine-tuning methods on a variety of tasks, making direct fine-tuning for quantized models possible. It therefore opens up the possibility for scaling up LLMs entirely in the quantized space. The source code is available at this https URL .
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2602.03120 [cs.LG]
(or arXiv:2602.03120v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2602.03120 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 272 words
Related coverage
The AI Inference Revolution Is Here
IEEE Spectrum · Sep 15, 2026
Selecting a vector store for Amazon Bedrock Knowledge Bases
AWS · Sep 17, 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · Sep 17, 2026
How to Size GPUs for AI Inference and TCO Without Overspending
NVIDIA · Sep 1, 2026