Krunoslav Lehman Pavasovic
@krunolp.bsky.social
PhD @Meta & @ENSParis. Diffusion Models, Flow Matching, High-Dimensional Statistics.
5/ The work is now available at:
📚 arXiv: arxiv.org/abs/2502.09445 #ICLR2025
💻GitHub: github.com/facebookrese...
📚 arXiv: arxiv.org/abs/2502.09445 #ICLR2025
💻GitHub: github.com/facebookrese...
GitHub - facebookresearch/difFOCI: GitHub repository for the work A Differentiable Rank-Based Objective for Better Feature Learning.
GitHub repository for the work A Differentiable Rank-Based Objective for Better Feature Learning. - facebookresearch/difFOCI
github.com
April 9, 2025 at 12:49 PM
5/ The work is now available at:
📚 arXiv: arxiv.org/abs/2502.09445 #ICLR2025
💻GitHub: github.com/facebookrese...
📚 arXiv: arxiv.org/abs/2502.09445 #ICLR2025
💻GitHub: github.com/facebookrese...
4/ On more complex machine-learning tasks, difFOCI performs favorably in terms of Worst group accuracy across several benchmarks:
April 9, 2025 at 12:49 PM
4/ On more complex machine-learning tasks, difFOCI performs favorably in terms of Worst group accuracy across several benchmarks:
3/ On standard classical feature selection tasks, difFOCI outperforms standard methods, selecting only a few informative, yet diverse features
April 9, 2025 at 12:49 PM
3/ On standard classical feature selection tasks, difFOCI outperforms standard methods, selecting only a few informative, yet diverse features
2/ difFOCI is a plug-and-play differentiable relaxation of the Chatterjee's correlation coefficient, a popular, recently-proposed rank-based estimator. We show that difFOCI improves on numerous ML and AI applications, including domain shift, spurious correlations and fairness
April 9, 2025 at 12:49 PM
2/ difFOCI is a plug-and-play differentiable relaxation of the Chatterjee's correlation coefficient, a popular, recently-proposed rank-based estimator. We show that difFOCI improves on numerous ML and AI applications, including domain shift, spurious correlations and fairness