Apple is presenting new research at the annual conference on Neural Information Processing Systems (NeurIPS), which takes place in-person in both Mexico City, Mexico, from November 30 – December 5, and San Diego, California, from December 2 – 7. We are proud to again sponsor the multi-track interdisciplinary conference, which brings together the scientific and industrial research communities surrounding Machine Learning. Below is an overview of Apple’s participation at NeurIPS 2025.

Jump to a section:

Stop by the Apple booth #1103 in the San Diego Convention Center during exhibition hours. All times listed in PST (San Diego local-time):

  • Tuesday, December 2, 12:00 PM - 8:00 PM
  • Wednesday, December 3, 9:00 AM - 5:00 PM
  • Thursday, December 4, 9:00 AM - 5:00 PM
  • Friday, December 5, 9:00 AM - 12:00 PM

Schedule

Sunday, November 30

Tuesday, December 2

Wednesday, December 3

Thursday, December 4

Friday, December 5

Saturday, December 6

Sunday, December 7

Technical Demos

MLX


We demonstrate large model inference and training on device using MLX. MLX is a flexible array framework that is optimized for Apple silicon and brought to you by Apple machine learning research. It enables training and inference of arbitrarily complex models on Apple silicon powered devices with great brevity and flexibility. We will showcase image generation using a large difusion model on an iPad Pro equipped with an M5 chip, and highlight disributed Compute with MLX and Apple Silicon. Visitors can explore text and code generation leveraging a 1 trillion parameter model running in Xcode on a 4-Mac Studio M3 Ultra cluster, each equipped with 512 GB of unified memory.

FastVLM


FastVLM is a family of mobile-friendly vision language models, built using MLX. These models use a mix of CNN and Transformer architectures for vision encoding designed specifically for processing high-resolution images. Together, they demonstrate a strong approach that achieves an optimal balance between accuracy and speed. Visitors can engage with FastVLM via a real-time visual question-and-answer demo on an iPhone 17 Pro Max.

Both MLX and FastVLM demos will be available at the Apple Booth during exhibition hours.

Main Venue-Accepted Papers

AuthorsChen Chen, Pengsheng Guo**, Liangchen Song, Jiasen Lu, Rui Qian, Xinze Wang, Tsu-Jui Fu**, Wei Liu**, Yinfei Yang, Alex Schwing
AuthorsVijay Viswanathan†, Yanchao Sun, Shuang Ma‡**, Xiang Kong, Meng Cao, Graham Neubig†, Tongshuang Wu†
AuthorsAlex Fang**†, Hadi Pouransari, Matt Jordan‡, Alexander Toshev, Vaishaal Shankar§**, Ludwig Schmidt†, Tom Gunter¶**
AuthorsMartin Pelikan*, Sheikh Shams Azam*, Vitaly Feldman, Jan “Honza” Silovsky, Kunal Talwar, Christopher G. Brinton†, Tatiana Likhomanenko*
AuthorsRuixiang Zhang, Shuangfei Zhai, Jiatao Gu, Yizhe Zhang, Huangjie Zheng, Tianrong Chen, Miguel Angel Bautista, Josh Susskind, Navdeep Jaitly
AuthorsLouis Béthune*, David Vigouroux†, Yilun Du‡, Rufin VanRullen§, Thomas Serre¶, Victor Boutin§*
AuthorsOrtal Senouf†, Antoine Wehenkel, Cédric Vincent-Cuaz†, Emmanuel Abbé, Pascal Frossard†
AuthorsPau Rodríguez‡, Michal Klein, Eleonora Gualdoni, Valentino Maiorca†**, Arno Blaas, Luca Zappella, Marco Cuturi, Xavier Suau‡

Models That Prove Their Own Correctness

Noga Amit (UC Berkeley), Shafi Goldwasser (UC Berkeley), Orr Paradise (UC Berkeley), Guy Rothblum

Sample and Map from a Single Convex Potential using Generative Moment Measures

Nina Vesseron (CREST-ENSAE), Louis Béthune, Marco Cuturi

AuthorsMustafa Shukor†, Louis Bethune, Dan Busbridge, David Grangier, Enrico Fini, Alaaeldin El-Nouby, Pierre Ablin
AuthorsJiatao Gu, Tianrong Chen, David Berthelot, Huangjie Zheng, Yuyang Wang, Ruixiang Zhang, Laurent Dinh, Miguel Angel Bautista, Josh Susskind, Shuangfei Zhai
AuthorsHaibo Wang‡‡, Bo Feng, Zhengfeng Lai, Mingze Xu, Shiyu Li, Weifeng Ge†, Afshin Dehghan, Meng Cao, Ping Huang
AuthorsRui Tian, Mingfei Gao, Mingze Xu, Jiaming Hu, Jiasen Lu, Zuxuan Wu†, Yinfei Yang, Afshin Dehghan

Acknowledgements

Samy Bengio is a Board Member for NeurIPS.

Pierre Ablin, Samy Bengio, Ronan Collobert, Marco Cuturi, Satyen Kale, Kunal Talwar, and Sinead Williamson are Senior Area Chairs.

Byeongjoo Ahn, Hilal Asi, Sheikh Shams Azam, Miguel Ángel Bautista, Meng Cao, Fartash Faghri, Mehrdad Farajtabar, Vitaly Feldman, Nick Foti, Zhe Gan, Jiatao Gu, Devon Hjelm, Chun-Liang Li, Tatiana Likhomanenko, Alex Papangelis, Stephan Richter, Pau Rodriguez, Skyler Seto, Natalie Schluter, Josh Susskind, Oncel Tuzel, Erik Wijmans, Yinfei Yang, Shuangfei Zhai, Yizhe Zhang, and Huangjie Zheng are Area Chairs.

Yannick Assogba, Deepika Bablani, Vassilis Baltatzis, Sima Behpour, Louis Bethune, Arno Blaas, Jie Bu, Chacha Chen, Chen Chen, Tianrong Chen, Masha Fedzechkina Donaldson, Fahad Kamran, Parmeshwar Khurd, Han-Byul Kim, Sunnie S. Y. Kim, Michael Kirchhof, Colin Lea, Pingzhi Li, Xintong Li, Ran Liu, Said Medjkouh, Lucas Monteiro Paes, Robin Schmidt, Amitis Shidani, Parshin Shojaee, Andrew Silva, Kaitao Song, Andrew Szot, Xinyu Tang, Vimal Thilak, William Thong, Raviteja Vemulapalli, Yuyang Wang, Dehong Xu, Jiayuan Ye, and Xiyou Zhou are Reviewers.

Fartash Faghri is a Workshop Co-Organizer; Samy Bengio, Mehrdad Farajtabar, Oncel Ouzel, and Hadi Pour Ansari serve on the Workshop Advisory Board; Mohammadreza Armandpour, Louis Béthune, Cheng-Yu Hsieh, Chen Huang, Cem Koc, Iman Mirzadeh and Raviteja Vemulapalli are Workshop Reviewers for Workshop on Continual and Compatible Foundation Model Updates (CCFM) 2025.

Antoine Wehenkel is a Workshop Co-Organizer for Machine Learning and the Physical Sciences Workshop 2025.

Related readings and updates.

Apple sponsored the Neural Information Processing Systems Conference (NeurIPS), which was held in New Orleans, Louisiana from November 28 to December 9. NeurIPS is a global conference focused on fostering the exchange of research on neural information processing systems in their biological, technological, mathematical, and theoretical aspects.

Read more

The Neural Information Processing Systems (NeurIPS) conference was held virtually on December 6 - 14. NeurIPS is a global conference focused on fostering the exchange of research on neural information processing systems in their biological, technological, mathematical, and theoretical aspects.

Read more