Aditya Kak
Electronics & VLSI Junior @ IIIT Delhi
New Delhi
1+ year of work experience
Jan 20269 mosPresent
Figuring out why LLMs work the way they do
Jun 20251 yr 4 mosPresent
Ann Arbor, Michigan, United States
Guide: Dr. Mehdi Saligane
Extending existing research on a LLM framework(https://dl.acm.org/doi/abs/
10.1145/3670474.3685971) for natural language to analog layout translation within
OpenFASOC(github.com/idea-fasoc/OpenFASOC) and gLayout (github.com/ReaLLMASIC/gLayout), employing fine-tuning techniques such as Instruction Fine-Tuning (IFT) and reinforcement learning-based post-training (GRPO etc.) to enhance analog circuit layout optimization and performance.
May 20243 mosJul 2024
Contributed to the SeekSuspect project, which uses LLMs to identify suspects from criminal dossiers based on visual memory. I helped train the LLaVA model to recognize facial features from text descriptions and developed an end-to-end pipeline using the LATS paper to digitally age suspect images. This approach enables the use of older photos for identifying suspects who may have aged or changed in appearance.
Aditya’s AI Stack
Education
Bachelor of Technology - BTech, Electronics & VLSI
Indraprastha Institute of Information Technology, Delhi
20234 yrs2027
High School Diploma
Springdales School, India
Join Aditya and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss