Research map / Federated learning and privacy / Personalized and heterogeneous FL
Heterogeneous Customizable Personalized Federated Fine-Tuning Approach for Large Language Models
xin tong, Baojiang cui
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 our scaling factor (personalization rank proportion) instead of none (ablation) | Het-CPFLoRA · Dataset 2 (heterogeneous) | Avg-Gen: 44.38 -> 47.49 | Improved |
| add our scaling factor instead of FedDPA's (ablation) | Het-CPFLoRA · Dataset 2 (heterogeneous) | Avg-Gen: 46.15 -> 47.49 | Improved |
| add our scaling factor instead of none (ablation) | Het-CPFLoRA · Dataset 2 (heterogeneous) | Avg-Per: 57.01 -> 58.91 | Improved |
| rear vs front position decoupling of the personalized subspace (ablation) | Het-CPFLoRA · Dataset 2 (rp=4, r=4, homogeneous) | Avg-Per / Avg-Gen: 52.27 -> 52.88 | No clear effect |
| increase personalized subspace rank share from 1:7 to 4:4 (ablation) | Het-CPFLoRA · Dataset 1 | Avg-Per: 54.22 -> 57.64 | Improved |
| increase personalized subspace rank share from 1:7 to 4:4 (ablation) | Het-CPFLoRA · Dataset 1 | Avg-Gen: 43.13 -> 38.63 | Hurt |
| add dynamic weighted inference (Our-auto vs Our) (ablation) | Het-CPFLoRA · Dataset 1 (All Homogeneous) | Avg-Gen: 38.48 -> 54.72 | Improved |
| add dynamic weighted inference (Our-auto vs Our) (ablation) | Het-CPFLoRA · Dataset 2 (All Homogeneous) | Avg-Per: 42.12 -> 52.27 | Improved |
| replace FedDPA's embedding-similarity weighting with our dynamic weighting (ablation) | Het-CPFLoRA · Dataset 1 (homogeneous) | Avg-Per: 56.77 -> 59.12 | Improved |
| use single-adapter decoupled LoRA (Our-auto) vs FedDPA-Auto dual adapter | Het-CPFLoRA · Dataset 2 (All Homogeneous) | Avg-Per: 45.86 -> 52.27 | Improved |
| scale to 48 clients with Het-CPFLoRA vs pFedSeq | Het-CPFLoRA · Dataset 1 (48 clients) | Avg-Per: 53.77 -> 59.36 | Improved |
| use Het-CPFLoRA on Qwen2.5-1.5B | Het-CPFLoRA · Dataset 2 (Qwen2.5-1.5B, heterogeneous) | Avg-Per: +6.64% | Improved |
Closest papers by meaning
- FLoRA: Federated Fine-Tuning Large Language Models with Heterogeneous Low-Rank Adaptations (NeurIPS 2024 · similarity 0.90)
- Adaptive LoRA Experts Allocation and Selection for Federated Fine-Tuning (NeurIPS 2025 · similarity 0.89)
- Federated Fine-tuning of Large Language Models under Heterogeneous Tasks and Client Resources (NeurIPS 2024 · similarity 0.89)
- Ravan: Multi-Head Low-Rank Adaptation for Federated Fine-Tuning (NeurIPS 2025 · similarity 0.88)
- Beyond Parameter Aggregation: Semantic Consensus for Federated Fine-Tuning of LLMs (NeurIPS 2026 · similarity 0.88)
- Robust Federated Finetuning of LLMs via Alternating Optimization of LoRA (NeurIPS 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.
