[
    {
        "id": "osp-17048",
        "type": "article-journal",
        "title": "HeuFouFT: Task-Guided Metaheuristic Coordinate Search for Fourier Fine-Tuning",
        "author": [
            {
                "family": "Wang",
                "given": "Ruiheng"
            },
            {
                "family": "Hou",
                "given": "Yubo"
            },
            {
                "family": "Zhu",
                "given": "Yakun"
            },
            {
                "family": "Shen",
                "given": "Tianle"
            },
            {
                "family": "Wan",
                "given": "Tao"
            },
            {
                "family": "Qin",
                "given": "Zengchang"
            }
        ],
        "URL": "https://omanscience.com/en/articles/heufouft-task-guided-metaheuristic-coordinate-search-for-fourier-fine-tuning",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "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."
    }
]