Hybrid Gated Flow (HGF): Stabilizing 1.58-bit LLMs via Selective Low-Rank Correction

• URL: http://arxiv.org/abs/2602.05269v1
• Date: Thu, 05 Feb 2026 03:47:17 GMT
• Title: Hybrid Gated Flow (HGF): Stabilizing 1.58-bit LLMs via Selective Low-Rank Correction
• Authors: David Alejandro Trejo Pizzo,
• Abstract summary: Hybrid Gated Flow (HGF) is a dual-stream architecture that couples a 1.58-bit ternary backbone with a learnable, low-rank FP16 correction path.<n>We show that HGF 5.4 achieves a validation loss of 0.9306 compared to BitNet's 1.0294, recovering approximately 55% of the quality gap between pure ternary quantization and the FP16 baseline.
• Score: 0.766310831583367
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The deployment of Large Language Models (LLMs) on edge devices is fundamentally constrained by the "Memory Wall" -- a hardware limitation where memory bandwidth, not compute, becomes the bottleneck. Recent 1.58-bit quantization techniques (e.g., BitNet b1.58) dramatically reduce memory footprint but typically incur a perplexity degradation of 20-25% compared to FP16 baselines. In this work, we introduce Hybrid Gated Flow (HGF), a dual-stream architecture that couples a 1.58-bit ternary backbone with a learnable, low-rank FP16 correction path controlled by adaptive gates. Through extensive experiments on the TinyStories dataset across two training regimes (2500 and 3500 steps), we demonstrate that HGF 5.4 achieves a validation loss of 0.9306 compared to BitNet's 1.0294, recovering approximately 55% of the quality gap between pure ternary quantization and the FP16 baseline (0.8490). This recovery is achieved with only ~12-15% memory overhead beyond the ternary backbone. Furthermore, we provide empirical evidence for an emergent phenomenon: quantization as structural regularization. While a full-precision differential attention baseline (Diff_Only) exhibited training instability with validation loss exceeding 1.68, the ternary-anchored HGF maintained robust convergence throughout training. Finally, we report preliminary results extending this architecture to 1.2B and 3B parameter models trained on SlimPajama and FineWeb-Edu. These larger-scale experiments confirm that the architectural stability and quality recovery observed in small-scale proxies scale linearly to production-grade language modeling regimes.

Related papers

• BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models [56.504879072674015]
  We propose Bit-Plane Decomposition Quantization (BPDQ), which constructs a variable quantization grid via bit-planes and scalar coefficients.<n>BPDQ enables serving Qwen2.5-72B on a single GTX 3090 with 83.85% GSM8K accuracy (vs. 90.83% at 16-bit)
  arXiv  Detail & Related papers  (2026-02-04T02:54:37Z)
• D3R-Net: Dual-Domain Denoising Reconstruction Network for Robust Industrial Anomaly Detection [0.0]
  Unsupervised anomaly detection (UAD) is a key ingredient of automated visual inspection in modern manufacturing.<n>We introduce D3R-Net, a Dual-Domain Denoising Reconstruction framework that couples a self-supervised 'healing' task with frequency-aware regularization.<n>In addition to the spatial mean squared error, we employ a Fast Fourier Transform (FFT) magnitude loss that encourages consistency in the frequency domain.
  arXiv  Detail & Related papers  (2026-01-27T23:21:59Z)
• Uncertainty Makes It Stable: Curiosity-Driven Quantized Mixture-of-Experts [6.221156050218661]
  We present a curiosity-driven quantized Mixture-of-Experts framework for deep neural networks on resource-constrained devices.<n>Our 4-bit quantization maintains 99.9 percent of 16-bit accuracy (0.858 vs 0.859 F1) with 4x compression and 41 percent energy savings.<n>Our information-theoretic routing demonstrates that adaptive quantization yields accurate (0.858 F1, 1.2M params), energy-efficient (3.87 F1/mJ), and predictable edge models.
  arXiv  Detail & Related papers  (2025-11-13T15:32:41Z)
• Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning [73.10669391954801]
  We present the Ring-linear model series, specifically including Ring-mini-linear-2.0 and Ring-flash-linear-2.0.<n>Both models adopt a hybrid architecture that effectively integrates linear attention and softmax attention.<n>Compared to a 32 billion parameter dense model, this series reduces inference cost to 1/10, and compared to the original Ring series, the cost is also reduced by over 50%.
  arXiv  Detail & Related papers  (2025-10-22T07:59:38Z)
• Ultrafast Deep Learning-Based Scatter Estimation in Cone-Beam Computed Tomography [7.864992877255044]
  scatter artifacts drastically degrade the image quality of cone-beam computed tomography (CBCT) scans.<n>Deep learning-based methods show promise in estimating scatter from CBCT measurements.<n>Their deployment in mobile CBCT systems or edge devices is still limited due to the large memory footprint of the networks.
  arXiv  Detail & Related papers  (2025-09-10T20:07:56Z)
• PRISM: Distributed Inference for Foundation Models at Edge [73.54372283220444]
  PRISM is a communication-efficient and compute-aware strategy for distributed Transformer inference on edge devices.<n>We evaluate PRISM on ViT, BERT, and GPT-2 across diverse datasets.
  arXiv  Detail & Related papers  (2025-07-16T11:25:03Z)
• SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer [49.1761733723771]
  This paper presents SANA-1.5, a linear Diffusion Transformer for efficient scaling in text-to-image generation.<n>We introduce three key innovations: Efficient Training Scaling, Model Depth Pruning, and Inference-time Scaling.<n>Through these strategies, SANA-1.5 achieves a text computation-image alignment score of 0.81 on GenEval, which can be further improved to 0.96 through inference scaling with VILA-Judge.
  arXiv  Detail & Related papers  (2025-01-30T15:31:48Z)
• Thinking Forward: Memory-Efficient Federated Finetuning of Language Models [21.438831528354513]
  Finetuning large language models (LLMs) in federated learning settings requires excessive memory for resource-constrained devices. In this paper, we introduce Spry, an FL algorithm that splits trainable weights of an LLM among participating clients. Spry achieves a low memory footprint, high accuracy, and fast convergence.
  arXiv  Detail & Related papers  (2024-05-24T13:37:48Z)
• Quantized Neural Networks for Low-Precision Accumulation with Guaranteed Overflow Avoidance [68.8204255655161]
  We introduce a quantization-aware training algorithm that guarantees avoiding numerical overflow when reducing the precision of accumulators during inference. We evaluate our algorithm across multiple quantized models that we train for different tasks, showing that our approach can reduce the precision of accumulators while maintaining model accuracy with respect to a floating-point baseline.
  arXiv  Detail & Related papers  (2023-01-31T02:46:57Z)
• Global Vision Transformer Pruning with Hessian-Aware Saliency [93.33895899995224]
  This work challenges the common design philosophy of the Vision Transformer (ViT) model with uniform dimension across all the stacked blocks in a model stage. We derive a novel Hessian-based structural pruning criteria comparable across all layers and structures, with latency-aware regularization for direct latency reduction. Performing iterative pruning on the DeiT-Base model leads to a new architecture family called NViT (Novel ViT), with a novel parameter that utilizes parameters more efficiently.
  arXiv  Detail & Related papers  (2021-10-10T18:04:59Z)
• Scale Efficiently: Insights from Pre-training and Fine-tuning Transformers [57.931830650323]
  This paper presents scaling insights from pretraining and finetuning Transformers. We show that aside from only the model size, model shape matters for downstream fine-tuning. We present improved scaling protocols whereby our redesigned models achieve similar downstream fine-tuning quality.
  arXiv  Detail & Related papers  (2021-09-22T12:29:15Z)
• Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks [9.409651543514615]
  This work introduces convolutional layers with pre-defined sparse 2D kernels that have support sets that repeat periodically within and across filters. Due to the efficient storage of our periodic sparse kernels, the parameter savings can translate into considerable improvements in energy efficiency.
  arXiv  Detail & Related papers  (2020-01-29T07:10:56Z)

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.