Scholar
Taha Ceritli
Google Scholar ID: 8qxNLQEAAAAJ
Samsung Research UK
Citations & Impact
Citations
128
H-index
6
i10-index
3
Publications
20
Co-authors
16
list available
Publications
Resume (English only)
Background
- Senior ML Researcher at Samsung Research UK, leading the Personalized AI team.
- Research interests include: adapting large language models (LLMs) in resource-constrained environments such as smartphones;
- parameter-efficient fine-tuning (PEFT) for efficient adaptation of LLMs to downstream tasks;
- decentralized training (e.g., federated learning) for privacy preservation;
- model merging to combine multiple LLMs or PEFT parameters for efficient deployment;
- continual learning for progressive model training over time;
- model compression techniques (e.g., quantization, pruning, knowledge distillation) to reduce LLM footprint;
- memory-based personalization of LLMs.
