WISCA: A Lightweight Model Transition Method to Improve LLM Training via Weight Scaling

arXiv:2508.16676v2 Announce Type: replace-cross Abstract: Transformer architecture gradually dominates the LLM field. Recent advances in training optimization for Transformer-based large language models (LLMs) primarily focus on architectural modifications or optimizer adjustments. However, these approaches lack systematic optimization of weight patterns during training. Weight pattern refers to the distribution and relative magnitudes of weight parameters in a neural network. To address this issue, we propose a Weight Scaling method called WISCA to enhance training efficiency and model quality by strategically improving neural network weight patterns without changing network structures. By rescaling weights while preserving model outputs, WISCA indirectly optimizes the model's training trajectory. Experiments demonstrate that WISCA significantly improves convergence quality (measured by generalization capability and loss reduction), particularly in LLMs with Grouped Query Attention (GQA) architectures and LoRA fine-tuning tasks. Empirical results show 5.6% average improvement on zero-shot validation tasks and 2.12% average reduction in training perplexity across multiple architectures.

Related

• Sparse or Dense? A Mechanistic Estimation of Computation Density in Transformer-based LLMs
• Dynamic Model Routing and Cascading for Efficient LLM Inference: A Survey
• Is More Data Worth the Cost? Dataset Scaling Laws in a Tiny Attention-Only Decoder
• A Mathematical Explanation of Transformers
• DySCO: Dynamic Attention-Scaling Decoding for Long-Context Language Models

Source: arXiv cs.CL | 2026-04-23