ReQAT: Achieving Full-Precision Reasoning Accuracy with 4-bit Floating-Point Quantization-Aware Training

• URL: http://arxiv.org/abs/2606.15682v1
• Date: Sun, 14 Jun 2026 09:00:06 GMT
• Title: ReQAT: Achieving Full-Precision Reasoning Accuracy with 4-bit Floating-Point Quantization-Aware Training
• Abstract summary: Large Reasoning Models (LRMs) achieve strong problem-solving through long chain-of-thought.<n>LRMs are constrained by the high cost of full-precision inference and growing KV cache footprints.<n>We propose ReQAT, a reasoning-centric FP4 training framework with three components.
• Score: 12.549469252572079
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Reasoning Models (LRMs) achieve strong problem-solving through long chain-of-thought, but their deployment is constrained by the high cost of full-precision inference and growing KV cache footprints. Microscaled FP4 formats enable efficient FP4 deployment; however, fully quantizing weights, activations, and KV caches (W4A4KV4) causes severe reasoning degradation that existing PTQ and QAT fail to recover. We identify that FP4 failures concentrate on low-entropy tokens--precise symbolic commitments such as digits and operators--where quantization noise inflates sampling errors that cascade through reasoning traces. Based on this insight, we propose ReQAT, a reasoning-centric FP4 training framework with three components: (i) Trace-Aligned QAT (TAQ), which revisits identical reasoning traces to focus updates on critical low-entropy decisions; (ii) Selective Entropy Minimization (SEM), which reinforces confidence at low-entropy positions; and (iii) Q-FIT, a quantization-friendly initialization that jointly calibrates RoPE-consistent KV cache transformations to stabilize QAT. Under the same training budget, ReQAT not only recovers but surpasses BF16 fine-tuning accuracy, while delivering up to 3.9x throughput speedup on NVIDIA DGX Spark and 3.1x on B200.

Related papers

• QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction [29.200349824129102]
  QUASAR continuously performs loss-aware reconstruction in the training loop to lower the loss floor and improve the resulting low-bit model.<n>It achieves the lowest held-out KL divergence among competitive QAT methods at 2, 3, and 4 bits.<n>It improves average accuracy across eight tasks by 3.5-4.3 percentage points over strong QAT and PTQ baselines.
  arXiv  Detail & Related papers  (2026-08-14T05:29:58Z)
• HiFloat4 Format for End-To-End Reinforcement Learning Post-Training of Large Language Models [9.149747042283197]
  We present the first end-to-end FP4 RL post-training, in which both the rollout and training policies operate at 4-bit precision.<n>A lightweight correction, Rollout-ResQ, recovers most of the precision lost to outlier-driven underflow without inflating the rollout's compute footprint.
  arXiv  Detail & Related papers  (2026-07-29T06:28:26Z)
• Full-Stack FP4: Stable LLM Pretraining with Quantized Projections, Optimizers, and Attention [34.07319488267708]
  Recent NVFP4 pretraining methods mainly target transformer linear layers, leaving computation states, arithmetic and attention underexplored in 4-bit pipelines.<n>This critical gap blocks stable full-stack 4-bit pretraining, as the three core modules exhibit unique numerical failure patterns.<n>We propose Full-Stack FP4, the first complete NVFP4 pretraining framework resolving all three stability bottlenecks.
  arXiv  Detail & Related papers  (2026-07-05T17:31:36Z)
• Attn-QAT: 4-Bit Attention With Quantization-Aware Training [10.003298547571447]
  This paper presents the first systematic study of 4-bit quantization-aware training (QAT) for attention.<n>We find that "drop-in" QAT, which naively combines an FP4 forward pass with a high-precision Flash Attention (FA)-style backward pass, leads to training instability.<n>We propose Attn-QAT and implement fused Triton kernels for training as well as FP4 inference kernels.
  arXiv  Detail & Related papers  (2026-02-09T04:46:21Z)
• A Case Study of Selected PTQ Baselines for Reasoning LLMs on Ascend NPU [7.030422837091069]
  Post-Training Quantization (PTQ) is crucial for efficient model deployment on Ascend NPU.<n>This paper presents a case study of PTQ baselines applied to reasoning-oriented models such as DeepSeek-R1-Distill-Qwen series (1.5B/7B/14B) and QwQ-32B.<n>We evaluate four distinct algorithms, including AWQ, GPTQ, SmoothQuant, and FlatQuant, to cover the spectrum from weight-only compression to advanced rotation-based methods.
  arXiv  Detail & Related papers  (2026-02-06T09:22:09Z)
• What Makes Low-Bit Quantization-Aware Training Work for Reasoning LLMs? A Systematic Study [59.44848132298657]
  Post-training quantization (PTQ) usually comes with the cost of large accuracy drops, especially for reasoning tasks under low-bit settings.<n>In this study, we present a systematic empirical study of quantization-aware training (QAT) for reasoning models.
  arXiv  Detail & Related papers  (2026-01-21T11:22:29Z)
• CAGE: Curvature-Aware Gradient Estimation For Accurate Quantization-Aware Training [73.46600457802693]
  We introduce a new method that counteracts the loss induced by quantization.<n>CAGE significantly improves upon the state-of-theart methods in terms of accuracy, for similar computational cost.<n>For QAT pre-training of Llama models, CAGE matches the accuracy achieved at 4-bits (W4A4) with the prior best method.
  arXiv  Detail & Related papers  (2025-10-21T16:33:57Z)
• Bridging the Gap Between Promise and Performance for Microscaling FP4 Quantization [77.67818998672516]
  We present the first comprehensive study of MXFP4 and NVFP4 for post-training quantization.<n>We introduce Micro-Rotated-GPTQ (MR-GPTQ), a variant of the classic GPTQ quantization algorithm.<n>We show that MR-GPTQ matches or outperforms state-of-the-art accuracy.
  arXiv  Detail & Related papers  (2025-09-27T09:22:21Z)
• End-to-End On-Device Quantization-Aware Training for LLMs at Inference Cost [53.25965863436039]
  Quantization-aware training (QAT) provides a more principled solution, but its reliance on backpropagation incurs prohibitive memory costs.<n>We propose ZeroQAT, a zeroth-order optimization-based QAT framework that supports both weight and activation quantization.<n>Experiments show that ZeroQAT consistently outperforms representative PTQ and QAT baselines while requiring significantly less memory.
  arXiv  Detail & Related papers  (2025-08-21T01:18:27Z)
• Pioneering 4-Bit FP Quantization for Diffusion Models: Mixup-Sign Quantization and Timestep-Aware Fine-Tuning [14.145862114439831]
  Model quantization reduces the bit-width of weights and activations, improving memory efficiency and inference speed.<n>Existing methods, primarily based on integer quantization and post-training quantization fine-tuning, struggle with inconsistent performance.<n>We propose a mixup-sign floating-point quantization framework, first introducing unsigned FP quantization in model quantization, along with timestep-aware LoRA and denoising-factor loss alignment.
  arXiv  Detail & Related papers  (2025-05-27T13:40:47Z)
• Optimizing Large Language Model Training Using FP4 Quantization [73.55459961002371]
  Quantized training presents a promising solution by enabling low-bit arithmetic operations to reduce costs.<n>This work introduces the first FP4 training framework for large language models (LLMs)
  arXiv  Detail & Related papers  (2025-01-28T18:04:50Z)
• GAQAT: gradient-adaptive quantization-aware training for domain generalization [54.31450550793485]
  We propose a novel Gradient-Adaptive Quantization-Aware Training (GAQAT) framework for DG.<n>Our approach begins by identifying the scale-gradient conflict problem in low-precision quantization.<n>Extensive experiments validate the effectiveness of the proposed GAQAT framework.
  arXiv  Detail & Related papers  (2024-12-07T06:07:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.