发表机构
Beihang University; Shanghai Jiao Tong University; VinUniversity（北京航空航天大学; 上海交通大学; VinUniversity）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
HeuFouFT提出任务引导的元启发式搜索（GA-SA、PSO、CS）选择傅里叶微调坐标，在E2E上优于现有方法，且FLOPs仅为全微调的15-18%。
AI 中文摘要
我们提出了启发式引导的傅里叶微调（HeuFouFT），这是一个任务引导的框架，用于在傅里叶微调中选择可训练的频率坐标。现有的均匀和高斯带通方案通过固定的、任务无关的规则分配有限的频谱预算。HeuFouFT 转而使用下游性能来搜索坐标。来自轻量级块级探测的粗略强度图初始化了三种元启发式优化器：带模拟退火的遗传算法（GA-SA）、粒子群优化（PSO）和布谷鸟搜索（CS）。在搜索过程中，随机森林过滤每个种群，使得只有前30%的候选者进入代理微调。在E2E数据集上使用GPT-2-Medium，所有三种变体在五个指标上均优于随机均匀傅里叶FT、高斯带通傅里叶FT和LoRA。PSO进一步在四个指标上优于最佳基线LoCA，同时使用的可训练频谱系数减少了37.6%。一旦坐标被选定，HeuFouFT仅需要全微调15-18%的FLOPs。这些结果表明，任务引导的搜索比固定采样更有效地分配有限的频谱容量。我们的代码已公开。
英文摘要
We introduce Heuristic-Guided Fourier Fine-Tuning (HeuFouFT), a task-guided framework for selecting trainable frequency coordinates in Fourier fine-tuning. Existing uniform and Gaussian band-pass schemes allocate a limited spectral budget through fixed, task-agnostic rules. HeuFouFT instead searches for coordinates using downstream performance. A coarse intensity map from lightweight block-level probes initializes three metaheuristic optimizers: Genetic Algorithm with Simulated Annealing (GA-SA), Particle Swarm Optimization (PSO), and Cuckoo Search (CS). During search, a Random Forest filters each population so that only the top 30% of candidates proceed to proxy fine-tuning. On E2E with GPT-2-Medium, all three variants outperform random-uniform FourierFT, Gaussian band-pass FourierFT, and LoRA across five metrics. PSO further outperforms LoCA, the best-performing baseline, on four metrics while using 37.6% fewer trainable spectral coefficients. Once coordinates are selected, HeuFouFT requires only 15--18% FLOPs of Full FT. These results show that task-guided search allocates limited spectral capacity more effectively than fixed sampling. Our code is publicly available.
