Research map / Federated learning and privacy / Personalized and heterogeneous FL
FDPT: Federated Discrete Prompt Tuning for Black-Box Visual-Language Models
Jiaqi Wu, Simin Chen, Jing Tang, Yuzhe Yang, Yiming Chen, Lixu Wang, Song Lin, Zehua Wang, Wei Chen, Zijian Tian
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add clustering strategy on server (ablation) | FDPT · Cifar-10 (Qwen2-VL-72B) | accuracy: 76.5 -> 78.5 | Improved |
| add clustering strategy on server (ablation) | FDPT · Cifar-100 (Qwen2-VL-72B) | accuracy: 70.6 -> 71.8 | Improved |
| use dynamic instead of fixed cluster count (ablation) | FDPT · Cifar-100 (Qwen2-VL-72B) | accuracy: 69.3 -> 71.8 | Improved |
| use weight comparison instead of random search for tokens (ablation) | FDPT · Cifar-10 (Qwen2-VL-72B) | accuracy: 65.3 -> 78.5 | Improved |
| use discrete token aggregation instead of embedding-fedavg (ablation) | FDPT · Cifar-10 (Qwen2-VL-72B) | accuracy: 71.9 -> 78.5 | Improved |
| add G-F mechanism on client (ablation) | FDPT · Cifar-10 (Qwen2-VL-72B) | accuracy: 71.8 -> 78.5 | Improved |
| add E-EB strategy on client (ablation) | FDPT · TinyImg (Qwen2-VL-72B) | accuracy: 56.2 -> 57.8 | Improved |
| add PPR-CoT strategy on client (ablation) | FDPT · Cifar-10 (Qwen2-VL-72B) | accuracy: 75.3 -> 78.5 | Improved |
| use discrete prompt tokens instead of soft prompts for client exchange (ablation) | FDPT · Communication overhead | parameters: 614k -> 150 | Improved |
| use detailed instead of shorter initial prompt (ablation) | all black-box methods · Cifar/TinyImg (GPT-4o) | accuracy | Improved |
| increase data heterogeneity (dir 10 to 0.05) (ablation) | FDPT · TinyImageNet heterogeneity (Qwen2-VL-72B) | accuracy: 57.1 -> 53.9 | Hurt |
| transfer prompts from InternVL2 to GPT4o | FDPT · Cifar-10 | accuracy: 79.0 -> 87.8 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FedBPT — FedBPT: Efficient Federated Black-box Prompt Tuning for Large Language Models (ICML 2024) · better on 4 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FedBPT: Efficient Federated Black-box Prompt Tuning for Large Language Models (ICML 2024 · similarity 0.90)
- Unlocking the Potential of Prompt-Tuning in Bridging Generalized and Personalized Federated Learning (CVPR 2024 · similarity 0.88)
- F^3OCUS - Federated Finetuning of Vision-Language Foundation Models with Optimal Client Layer Updating Strategy via Multi-objective Meta-Heuristics (CVPR 2025 · similarity 0.88)
- Global and Local Prompts Cooperation via Optimal Transport for Federated Learning (CVPR 2024 · similarity 0.88)
- FedPHA: Federated Prompt Learning for Heterogeneous Client Adaptation (ICML 2025 · similarity 0.88)
- FOCoOp: Enhancing Out-of-Distribution Robustness in Federated Prompt Learning for Vision-Language Models (ICML 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
