Скачать или смотреть AI Frontiers: LLM Safety, RL, Diffusion Models & More | Sept 26, 2025

Welcome to AI Frontiers! This week, we dive into 16 groundbreaking papers from the cs.LG (Machine Learning) arXiv category, released on September 26th, 2025. Our AI-powered synthesis, using models like Gemini 2.5 Flash Lite and curated by GPT, highlights key advancements shaping the future of artificial intelligence.

1. Jiayin Liu et al. (2025). Physically Plausible Multi-System Trajectory Generation and Symmetry Discovery. http://arxiv.org/pdf/2509.23003v1

2. Konstantina Bairaktari et al. (2025). Sample-efficient Multiclass Calibration under $\ell_{p}$ Error. http://arxiv.org/pdf/2509.23000v1

3. Zachary Baker et al. (2025). Analysis of Variational Autoencoders. http://arxiv.org/pdf/2509.22994v1

4. Yuanyuan Yang et al. (2025). T-TAMER: Provably Taming Trade-offs in ML Serving. http://arxiv.org/pdf/2509.22992v1

5. David P. Morton et al. (2025). MDP modeling for multi-stage stochastic programs. http://arxiv.org/pdf/2509.22981v1

6. Zeyi Chen et al. (2025). OptiMind: Teaching LLMs to Think Like Optimization Experts. http://arxiv.org/pdf/2509.22979v1

7. Samuel V. Singh et al. (2025). Shape-Informed Clustering of Multi-Dimensional Functional Data via Deep Functional Autoencoders. http://arxiv.org/pdf/2509.22969v1

8. Qinxun Bai et al. (2025). Functional Critic Modeling for Provably Convergent Off-Policy Actor-Critic. http://arxiv.org/pdf/2509.22964v1

9. Haitong Ma et al. (2025). Reinforcement Learning with Discrete Diffusion Policies for Combinatorial Action Spaces. http://arxiv.org/pdf/2509.22963v1

10. Luke Guerdan et al. (2025). Doubly-Robust LLM-as-a-Judge: Externally Valid Estimation with Imperfect Personas. http://arxiv.org/pdf/2509.22957v1

11. Valentyn Melnychuk et al. (2025). GDR-learners: Orthogonal Learning of Generative Models for Potential Outcomes. http://arxiv.org/pdf/2509.22953v1

12. Hamidreza Moazzami et al. (2025). Meta-Learning Fourier Neural Operators for Hessian Inversion and Enhanced Variational Data Assimilation. http://arxiv.org/pdf/2509.22949v1

13. Lorenz K. Müller et al. (2025). SINQ: Sinkhorn-Normalized Quantization for Calibration-Free Low-Precision LLM Weights. http://arxiv.org/pdf/2509.22944v1

14. Yanqing Lu et al. (2025). Understanding SOAP from the Perspective of Gradient Whitening. http://arxiv.org/pdf/2509.22938v1

15. Aleksandr Dremov et al. (2025). Compute-Optimal Quantization-Aware Training. http://arxiv.org/pdf/2509.22935v1

16. Shreyas Gokhale (2025). MonoCon: A general framework for learning ultra-compact high-fidelity representations using monotonicity constraints. http://arxiv.org/pdf/2509.22931v1

17. Matthieu Zimmer et al. (2025). Rethinking Large Language Model Distillation: A Constrained Markov Decision Process Perspective. http://arxiv.org/pdf/2509.22921v1

18. Jake S. Rhodes et al. (2025). Guided Manifold Alignment with Geometry-Regularized Twin Autoencoders. http://arxiv.org/pdf/2509.22913v1

19. Anutam Srinivasan et al. (2025). FedCF: Fair Federated Conformal Prediction. http://arxiv.org/pdf/2509.22907v1

20. Karim Khamaisi et al. (2025). From Noise to Knowledge: A Comparative Study of Acoustic Anomaly Detection Models in Pumped-storage Hydropower Plants. http://arxiv.org/pdf/2509.22881v1

Disclaimer: This video uses arXiv.org content under its API Terms of Use; AI Frontiers is not affiliated with or endorsed by arXiv.org.