Tootfinder

Unsupervised Incremental Learning Using Confidence-Based Pseudo-Labels
Lucas Rakotoarivony
https://arxiv.org/abs/2508.21424 https://arxiv.org/pdf/2508.2142…

Unsupervised Incremental Learning Using Confidence-Based Pseudo-Labels
Deep learning models have achieved state-of-the-art performance in many computer vision tasks. However, in real-world scenarios, novel classes that were unseen during training often emerge, requiring models to acquire new knowledge incrementally. Class-Incremental Learning (CIL) methods enable a model to learn novel classes while retaining knowledge of previous classes. However, these methods make the strong assumption that the incremental dataset is fully labeled, which is unrealistic in pract…

Simulating Unruh Radiation in High-Intensity Laser-Electron Interactions for Near-Term Experimental Tests
Rafi Hessami, Haidar Al-Naseri, Monika Yadav, Maanas Hemanth Oruganti, Brian Naranjo, James Rosenzweig
https://arxiv.org/abs/2509.07386

Simulating Unruh Radiation in High-Intensity Laser-Electron Interactions for Near-Term Experimental Tests
The Unruh effect predicts that a uniformly accelerating observer perceives the vacuum as a thermal bath, yet direct observation remains elusive [1]. We simulate Unruh radiation in realistic high-intensity laser-electron collisions relevant to FACET-II and LUXE using fully three-dimensional Monte Carlo methods. In our model, Unruh emission is treated as scattering from a rest-frame thermal spectrum with Klein-Nishina cross sections, while nonlinear Compton radiation is computed across many harmo…