Zhyper：用于条件化LLM微调的因子化超网络
Zhyper: Factorized Hypernetworks for Conditioned LLM Fine-Tuning
AI总结：
提出参数高效的因子化超网络框架Zhyper，从文本描述生成上下文感知LoRA适配器，以少26倍参数实现有竞争力性能，并扩展至文化对齐提升泛化。
摘要：
大语言模型（LLM）条件化是指指示LLM根据特定文化的规范和价值观、特定政治倾向的信念或任何期望的文本指定语义条件来生成内容。不幸的是，由于预训练和对齐数据集的归纳偏差，提示工程并不能确保LLM按照期望的条件化行为行事。先前的工作侧重于通过直接条件化LoRA权重来微调LLM；然而，此类方法引入了大量参数。作为解决方案，我们提出了Zhyper，一种参数高效的因子化超网络框架，它从文本描述中生成上下文感知的LoRA适配器。在多个基准上的实验表明，Zhyper以比最先进的基线少多达26倍的参数实现了有竞争力的性能。此外，我们将Zhyper扩展到文化对齐，展示了在域外设置中改进的泛化能力以及对细粒度上下文价值观的更好捕捉。
英文摘要：
Large Language Model (LLM) conditioning refers to instructing an LLM to generate content in accordance with the norms and values of a specific culture, beliefs of a particular political orientation, or any desired text-specified semantic conditioning. Unfortunately, prompt engineering does not ensure that LLMs behave in accordance with a desired conditioning due to the inductive bias of the pre-training and alignment datasets. Prior works have focused on fine-tuning LLMs by directly conditioning the LoRA weights; however, such methods introduce a large number of parameters. As a remedy, we propose Zhyper, a parameter-efficient factorized hypernetwork framework that generates context-aware LoRA adapters from textual descriptions. Experiments on multiple benchmarks show that Zhyper achieves competitive performance with up to 26x fewer parameters than the state-of-the-art baselines. Furthermore, we extend Zhyper to cultural alignment, demonstrating improved generalization to out-of-domain settings and a better capturing of fine-grained contextual values.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
