Hello! I'm Alper,
I'm a research engineer, worked on multimodal models, llms, diffusion models, stt/tts for low-resource languages and domains (Turkish, telephony, etc). My background is in mathematics, thus I'm really interested in math-heavy part of the deep learning too. I like to practice interpretability and efficiency in LLMs and generative models. I'm particularly interested in transformer-based models, diffusion models, and mechanistic interpretability. Currently I'm interested in:
- Mechanistic interpretability of generative models
- Efficiency: pruning, distillation, quantization for LLMs and diffusion
- Architecture-level analysis of transformers and diffusion models
- GPU systems and inference optimization
- Low-resource multilingual speech modeling (as prior specialty)
