用微调参数控制大模型输出方向，省算力还保效果。
Propulsion: Steering LLM with Tiny Fine-Tuning
- 通过可训练小参数重标定模型维度，不改原参数
- 参数量从355.3万降至0.086万，降超10倍
- 适合资源有限却需高效定制大模型的场景
大型语言模型（LLM）的快速发展推动了自然语言处理的进步，但针对特定任务的微调仍面临计算开销大、易破坏预训练知识的问题。为此，我们提出Propulsion，一种新型参数高效微调（PEFT）方法，旨在优化任务性能的同时大幅降低计算成本。受物理运动中可控调节的启发，Propulsion仅选择性地重新缩放预训练模型的特定维度，引导输出预测向任务目标靠拢，而无需修改模型参数。通过在预训练层引入轻量级可训练的Propulsion参数，显著减少微调时需更新的参数数量，防止过拟合或覆盖已有知识。理论分析结合神经正切核（NTK）理论表明，Propulsion能以极少可训练参数逼近全量微调的表现。实验上，参数量由355.3百万降至0.086百万，相比标准方法（如LoRA）实现超10倍压缩，且在多个基准测试中保持竞争力。
原文摘要 · Abstract (English)
The rapid advancements in Large Language Models (LLMs) have revolutionized natural language processing (NLP) and related fields. However, fine-tuning these models for specific tasks remains computationally expensive and risks degrading pre-learned features. To address these challenges, we propose Propulsion, a novel parameter efficient fine-tuning (PEFT) method designed to optimize task-specific performance while drastically reducing computational overhead. Inspired by the concept of controlled adjustments in physical motion, Propulsion selectively re-scales specific dimensions of a pre-trained model, guiding output predictions toward task objectives without modifying the model's parameters. By introducing lightweight, trainable Propulsion parameters at the pre-trained layer, we minimize the number of parameters updated during fine-tuning, preventing overfitting or overwriting of existing knowledge. Our theoretical analysis, supported by Neural Tangent Kernel (NTK) theory, shows that Propulsion approximates the performance of full fine-tuning with far fewer trainable parameters. Empirically, Propulsion reduces the parameter count from 355.3 million to just 0.086 million, achieving over a 10x reduction compared to standard approaches like LoRA while maintaining competitive performance across benchmarks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。