Links My CV (PDF) Google Scholar
About me
Welcome to my home page! I am a PhD student at McGill University, supervised by Prof. Mark Coates. My thesis is on automated logical reasoning. In particular, I am interested in modelling human reasoning processes as probabilistic, sequential processes.
Publications
2026
- Commit Early, Refine Later? Probing Recursive Reasoning in TRM
D. Karakus, J. Cotnareanu, M. Coates.
Proc. Neural Network Artifacts as a New Data Modality Workshop at NeurIPS, 2026.
[PDF] - DynaPFN: Zero-Shot Dynamical System Forecasting with Tabular Prior-Fitted Networks
C. Roverato, J. Cotnareanu, et al.
Advances in Neural Information Processing Systems (NeurIPS), 2026.
[PDF] - Is the acquisition worth the cost? Surrogate losses for Consistent Two-stage Classifiers
F. Regol, J. Cotnareanu, et al.
Advances in Neural Information Processing Systems (NeurIPS), 2026. Spotlight.
[PDF] - Abductive Reasoning with Probabilistic Commonsense
J. Cotnareanu, C. Roverato, et al.
Proc. Int. Conf. Machine Learning (ICML), 2026.
[PDF] - A Balanced Neurosymbolic Approach to Abductive Commonsense Reasoning
J. Cotnareanu, D. Chetelat, et al.
Proc. Int. Conf. Learning Representations (ICLR), 2026.
[PDF]
2025
- InnerThoughts: Disentangling Representations and Predictions in Large Language Models
D. Chetelat, J. Cotnareanu, et al.
Proc. Int. Conf. Artificial Intelligence and Statistics (AISTATS), 2025.
[PDF]
2024
- HardCore Generation: Generating Hard UNSAT Problems for Data Augmentation
J. Cotnareanu, Z. Zhang, et al.
Advances in Neural Information Processing Systems (NeurIPS), 2024.
[PDF] - GRASS: Combining Graph Neural Networks with Expert Knowledge for SAT Solver Selection
Z. Zhang, D. Chetelat, J. Cotnareanu, et al.
Proc. Conf. Knowledge Discovery and Data Mining (KDD), 2024.
[PDF]
2023
- Substituting Data Annotation with Balanced Updates and Collective Loss in Multi-label Text Classification
M. Ozmen, J. Cotnareanu.
Conf. Lifelong Learning Agents (CoLLAs), 2023.
[PDF]