Research map / Federated learning and privacy / Privacy attacks and secure learning
PRISM: Privacy-Aware Routing for Adaptive Cloud–Edge LLM Inference via Semantic Sketch Collaboration
Junfei Zhan, Haoxun Shen, Zheng Lin, Tengjiao He
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 |
|---|---|---|---|
| use PRISM adaptive routing instead of Uniform LDP | Cloud-edge LDP inference · Semi-synthetic privacy prompts (4 domains) | completion time (s): 20.56 -> 7.92 | Improved |
| use PRISM instead of Uniform LDP | Cloud-edge LDP inference · Semi-synthetic privacy prompts (4 domains) | edge energy (J): 1707.6 -> 687.16 | Improved |
| use PRISM instead of Uniform LDP | Cloud-edge LDP inference · Semi-synthetic privacy prompts (4 domains) | inference quality: 5.72 -> 6.88 | Improved |
| use PRISM instead of Selective LDP | Cloud-edge LDP inference · Semi-synthetic privacy prompts (4 domains) | inference quality: 5.94 -> 6.88 | Improved |
| use PRISM instead of Edge-Only inference | SLM inference · Semi-synthetic privacy prompts (4 domains) | inference quality: 5.09 -> 6.88 | Improved |
| use PRISM instead of Cloud-Only inference (privacy gained) | LLM inference · Semi-synthetic privacy prompts (4 domains) | inference quality: 8.14 -> 6.88 | Hurt |
| use PRISM instead of Cloud-Only inference | LLM inference · Semi-synthetic privacy prompts (4 domains) | edge energy (J): 296.27 -> 687.16 | Hurt |
| pair GPT-4o with TinyLLaMA-1.1B edge model instead of StableLM-2-Zephyr-1.6B (ablation) | PRISM · Semi-synthetic privacy prompts (4 domains) | inference quality: 7.16 -> 5.28 | Hurt |
| vary privacy budget epsilon (ablation) | PRISM · Semi-synthetic privacy prompts (4 domains) | inference quality | No clear effect |
Closest papers by meaning
- HybridFlow: Resource-Adaptive Subtask Routing for Efficient Edge-Cloud LLM Inference (ICML 2026 · similarity 0.87)
- Cape: Context-Aware Prompt Perturbation Mechanism with Differential Privacy (ICML 2025 · similarity 0.86)
- Portcullis: A Scalable and Verifiable Privacy Gateway for Third-Party LLM Inference (AAAI 2025 · similarity 0.86)
- Split-and-Denoise: Protect large language model inference with local differential privacy (ICML 2024 · similarity 0.86)
- Hidden No More: Attacking and Defending Private Third-Party LLM Inference (ICML 2025 · similarity 0.86)
- Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory (ICLR 2024 · similarity 0.85)
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.
