PAE MobiLLM combines additive side-tuning, token-selected activation transmission, and server-side activation caching to make on-device LLM fine-tuning faster and cheaper, while transmitting prediction differences instead of raw labels.
In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
PAE MobiLLM: Privacy-Aware and Efficient LLM Fine-Tuning on the Mobile Device via Additive Side-Tuning
PAE MobiLLM combines additive side-tuning, token-selected activation transmission, and server-side activation caching to make on-device LLM fine-tuning faster and cheaper, while transmitting prediction differences instead of raw labels.