# Towards Efficient Pre-training: Exploring FP4 Precision in Large Language Models

arXiv: https://arxiv.org/abs/2502.11458 · HTML: https://mlsanity.com/paper/2502.11458/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Feb 2025.

## Built on by (most cited first, 8 of 8)

- [Pretraining Large Language Models with NVFP4](https://mlsanity.com/paper/2509.25149/) (2025)
- [Low-Precision Training of Large Language Models: Methods, Challenges, and Opportunities](https://mlsanity.com/paper/2505.01043/) (2025)
- [Pretraining large language models with MXFP4 on Native FP4 Hardware](https://mlsanity.com/paper/2605.09825/) (2026)
- [Elucidating the Design Space of FP4 training](https://mlsanity.com/paper/2509.17791/) (2025)
- [Stable FP4 Training via Transposition-Invariant Block Quantization](https://mlsanity.com/paper/2607.24953/) (2026)
- [UE5M3 FP4 Block Scaling for Stable Language Model Pretraining](https://mlsanity.com/paper/2609.02846/) (2026)
- [Performance of Small Language Model Pretraining on FABRIC: An Empirical Study](https://mlsanity.com/paper/2602.02632/) (2026)
- [Format-Aware Fusion for Fast FP4 Pretraining](https://mlsanity.com/paper/2610.00053/) (2026)

## Abstract (arXiv)

The burgeoning computational demands for training large language models (LLMs) necessitate efficient methods, including quantized training, which leverages low-bit arithmetic operations to reduce costs. While FP8 precision has shown potential, leveraging FP4 remains challenging due to inherent quantization errors and limited representation capability. Based on the Transformer architecture, we present an FP4 training scheme for LLMs, overcoming these obstacles through mixed-precision quantization strategies tailed for different modules and training stages. This allows us to apply the precision level suitable to distinct components within the model, ensuring that multi-head attention and linear layers are handled appropriately. Our pretraining recipe ensures stability in backpropagation by incorporating fine-grained quantization methods with a target precision training schedule. Experimental results demonstrate that our FP4 training scheme achieves accuracy comparable to BF16 and FP8, with smaller theoretical computational cost. With the advent of next-generation hardware supporting FP4, our method sets the foundation for efficient ultra-low precision training.