I am a Master's student at the University of Illinois at Urbana-Champaign (UIUC) advised by Prof. Tong Zhang. I am interested in improving and understanding how we can not only help models learn more efficiently but also align them with our goals and values. Methods I am currently researching include continual learning, curriculum learning, and memory structures.
Previously, I worked as a quantitative trader at Optiver US LLC and Old Mission Capital, and as a software engineer at VectorShift, Amazon, and Secureframe.
I am also a classically trained pianist (enjoy).
") does not match the recommended repository name for your site ("
").
", so that your site can be accessed directly at "http://
".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}
" in index.html
.
",
which does not match the baseurl
("
") configured in _config.yml
.
baseurl
in _config.yml
to "
".
Jerry Huang, Siddarth Madala, Risham Sidhu, Cheng Niu, Hao Peng, Julia Hockenmaier, Tong Zhang
Under review. 2025
We introduce RAG-RL, an answer generation model trained to identify and reason over larger sets of retrieved information using reinforcement learning and curriculum learning.
Jerry Huang, Siddarth Madala, Risham Sidhu, Cheng Niu, Hao Peng, Julia Hockenmaier, Tong Zhang
Under review. 2025
We introduce RAG-RL, an answer generation model trained to identify and reason over larger sets of retrieved information using reinforcement learning and curriculum learning.
For my undergraduate thesis, I investigated the feasibility of using software to generate fingerprint lineups (similar eye-witness lineups) in order to quantify the amount of cognitive bias present in the field of forensic science.
For my undergraduate thesis, I investigated the feasibility of using software to generate fingerprint lineups (similar eye-witness lineups) in order to quantify the amount of cognitive bias present in the field of forensic science.