Scholar
George Tsoukalas
Google Scholar ID: XbfbI7oAAAAJ
PhD Student, UT Austin
Citations & Impact
Citations
104
H-index
3
i10-index
2
Publications
5
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- The paper 'PutnamBench' was accepted at NeurIPS 2024 and won the Best Paper Award at the ICML 2024 AI for Math Workshop; contributed to COPRA, an agentic LLM-based approach for formal theorem-proving; conducted research on differentiable programming for invariant synthesis for program verification at the Rutgers Automated Reasoning Lab; participated in an NSF REU project, studying eigenvalue assignment of certain graphs, and submitted a paper based on the results; studied symmetric numerical semigroups at San Diego State University, advised by Chris O’Neill, and characterized the facets of unimaximal faces of the Kunz Cone.
Research Experience
- In his second year at UT Austin, he has been involved in two projects broadly falling under the umbrella term “AI for Mathematics.” He organized a team to produce a new formal competition maths benchmark, PutnamBench, which was accepted at NeurIPS 2024 and recognized at the ICML 2024 AI for Math Workshop via the Best Paper Award. He also contributed to COPRA, an agentic LLM-based approach for formal theorem-proving.
Education
- Ph.D. student in Computer Science at the University of Texas at Austin, advised by Dr. Swarat Chaudhuri; conducted research at the Rutgers Automated Reasoning Lab as an undergraduate, advised by Prof. He Zhu; participated in an NSF REU at the College of William & Mary, advised by Prof. Charles R. Johnson.
Background
- A second-year Ph.D. student in Computer Science at the University of Texas at Austin, with research interests in neurosymbolic techniques for automated mathematical reasoning and program synthesis.
Miscellany
- Personal website built using Jekyll and AcademicPages.
Co-authors