Disordered Systems & Neural Networks

Lorenzo Bardone, Claudia Merger, Sebastian Goldt

This provides foundational insights into how diffusion models learn complex distributions, which is incremental but clarifies a key mechanism for the AI/ML community.

Kaito Takanami, Cengiz Pehlevan

Provides a unified theoretical understanding of how test-time reasoning depth affects generalization in LLMs, addressing a poorly understood scaling behavior.

Clarissa Lauditi, Cengiz Pehlevan, Blake Bordelon

Provides a theoretical framework for understanding feature learning and hyperparameter transfer in wide neural networks, relevant to practitioners scaling up models.

Jiawei Guo, Daniel Schwalbe-Koda

This addresses the intricate challenge of structure elucidation in amorphous materials for materials scientists, offering a novel method that bypasses expert guidance or potentials, though it builds on existing generative techniques.

Louie Hong Yao, Yuhao Li, Shengchao Liu

For researchers studying self-supervised learning, this work provides a theoretical understanding of collapse mechanisms and prevention, though the model is highly simplified.

Luca Maria Del Bono, Giulio Biroli, Patrick Charbonneau et al.

Provides theoretical insight into the limitations of diffusion models near criticality and demonstrates how architectural design can overcome these bottlenecks, relevant for statistical physics and generative modeling.

Enrico Ventura, Beatrice Achilli, Luca Ambrogioni et al.

For practitioners using CFG in diffusion models, this work provides a theoretical understanding of diversity loss and a practical fix, though the analysis is limited to Gaussian mixtures.

Fabiola Ricci, Claudia Merger, Sebastian Goldt

For researchers studying learning dynamics and generalization in neural networks, this work provides a theoretical and experimental framework linking Fourier properties of data to sample complexity and training speed.

Taichi Haruna, Kohei Nakajima

This work provides a theoretical framework for understanding memory limits in recurrent neural networks, relevant to reservoir computing and neuromorphic systems, though the results are primarily analytical with limited immediate practical impact.

Ruomin Zhu, Abhishek Kumar Singh, Jérémie Laydevant et al.

This work solves a key bottleneck in probabilistic Ising machines—parallel updates—enabling faster solvers for dense optimization problems, with demonstrated utility in real-time wireless communications.

John Sous, Michael Winer

This work provides a theoretical framework for understanding scaling laws in sparse neural networks, which is important for practitioners designing efficient models under compute constraints.

Ejaaz Merali, Mohamed Hibat-Allah, Mohammad Kohandel et al.

This work makes recurrent neural network quantum states practical for large-scale quantum many-body simulations, addressing a scalability bottleneck.

Antoine Maillard, Sebastian Goldt

This work clarifies the fundamental distinction between convergence and latent factor recovery in generative models, providing theoretical insights for practitioners regarding data requirements and evaluation metrics.

Shu Zhou, K. Y. Michael Wong, Juntao Wang et al.

This work addresses the lack of theoretical insights for practitioners using analog solvers, though it is incremental as it builds on existing dynamical systems approaches.

Hidenori Tanaka

This provides a framework for understanding social representation formation in multi-agent systems, with implications for AI deployment in decision-making, though it is incremental as it builds on prior naming-game studies.

Alessio Giorlandino, Sebastian Goldt, Antoine Maillard

This work establishes a fundamental baseline for understanding memory capacity in neural networks, relevant to researchers studying factual recall in large language models.

Jie Huang, Bruno Loureiro, Stefano Sarao Mannelli

This provides foundational insights into optimization challenges in neural networks, addressing a core problem for machine learning researchers.

Brice Huang, Mark Sellke

This work provides theoretical evidence for computational hardness in spin glass optimization, impacting physics and algorithm design, though it is incremental on prior conjectures.

Gen Zu, Ning Mao, Claudia Felser et al.

This addresses the need for transparent, chemistry-based materials informatics to generate scientific insights beyond black-box predictions, representing a new paradigm rather than an incremental improvement.

M Mahmudul Hasan Sajeeb, Kevin Callahan-Coray, Corentin Delacour et al.

This work addresses the dense connectivity bottleneck in probabilistic computers for real-world combinatorial optimization, offering a tuning-free algorithmic framework that could enable scalable hardware for next-generation wireless MIMO detection.