Split-and-Scale Latent 3D Representations
Researchers Dorian Chan, Muhammed Kocabas, Xiaoming Zhao, Oncel Tuzel and Jen-Hao Rick Chang published a paper on split-and-scale, an inference-time method that expands the latent capacity of 3D repre…
Researchers Dorian Chan, Muhammed Kocabas, Xiaoming Zhao, Oncel Tuzel and Jen-Hao Rick Chang published a paper on split-and-scale, an inference-time method that expands the latent capacity of 3D repre…
A November 2026 paper by Lucas Monteiro Paes, Natalie Mackraz, Barry-John Theobald and Federico Danieli derives a closed-form expression for the maximum achievable expected reward gain under a fixed K…
A NeurIPS 2026 workshop paper from researchers at the University of Southern California and the University of Copenhagen finds that LLM judges' quality correlates only weakly with general model abilit…
Researchers William L. Tong, Aryo Lotfi, Emmanuel Abbe, Josh Susskind, Cengiz Pehlevan and Eran Malach proposed weighted state souping, a method that learns query-dependent weights for composing cache…
Apple researchers Audra McMillan, Kunal Talwar and Felix Zhou published a paper in November 2026 recasting Private Evolution (PE), a differentially private synthetic data generation algorithm, as gene…
A new research paper, "Toward a Theory of Evaluability," formalizes how black-box machine learning models can be evaluated when test samples may have been used during training, dropping the classical …
Researchers from Apple presented a paper at ICML 2026, "Amortizing Maximum Inner Product Search with Learned Support Functions," that trains small neural networks offline for a fixed database of keys …
Apple's MLX framework, in a release published today, runs the 816-billion-parameter Muse Spark Open model at over 36 tokens per second on a cluster of four Mac Studios connected via the RDMA over Thun…
A scale study by Mingfei Zhang, Amit Surana, Ziauddin Majid, Chenxi Wang and Michael Xu found a complexity-dependent crossover between two tool-discovery architectures for LLM agents: Retrieve-then-Pl…
Researchers from the University of Cambridge, MIT, Google DeepMind and Apple introduced the first tri-modal Masked Diffusion Models pretrained from scratch on text, image-text and audio-text data, pre…
A paper published in October 2026 by Jianhua Li, Yun Xing, Girisha De Silva, Ryan Flint, Kai Yuan and Sean Suchter presents Minions, a dynamic and distributed virtual GPU cloud that aggregates idle Ap…
Johns Hopkins University researchers Tianjian Li, Jingyu Zhang, William Jurayj, Xi Wang, Chuanyang Jin, Mehrdad Farajtabar, Eric Nalisnick and Daniel Khashabi published SELFCOMPACT in October 2026, a …
Apple researchers Jiatao Gu, Tianrong Chen, Ying Shen, David Berthelot, Shuangfei Zhai and Josh Susskind published Normalizing Trajectory Models (NTM) in October 2026, a generative modeling method tha…
Researchers Jingtan Wang, Sirajul Salekin, Young mok Jung, Javier Movellan, Bryan Kian Hsiang Low and Manjot Bilkhu published RISED, a method that repurposes rubrics — textual descriptions of rollout …
Apple researchers published a paper in October 2026 titled "Negotiating Ontological Boundaries in User-Authored Personal Sensing Systems," reporting a week-long exploratory study in which participants…
Apple researchers Russ Webb, Amitis Shidani, Alice Bizeul and Dan Busbridge published an October 2026 paper showing that discrete diffusion samplers — including remasking and uniform-state samplers — …
A paper published in October 2026 by Kirill Brilliantov, Alejandro Hernández-Cano and Emmanuel Abbé finds that open-source state-of-the-art MLE agent harnesses provide no advantage over a single sessi…
A new method called RLTL;DR lets a Qwen 3.5 9B Thinking policy break through a learning barrier on tool-calling and coding datasets filtered to Pass@128 = 0, reaching a Pass@1 of 14–31% with insights …
A systematic study by Iuri Macocco, Pau Rodríguez Lopez, Arno Blaas, Luca Zappella, Marco Baroni and Xavier Suau Cuadros, published in Transactions on Machine Learning Research (TMLR) in September 202…
A round-trip study of sixteen language models found that serializing tree-structured arithmetic expressions into natural language is a lossy, asymmetric channel, with swapping the generating and extra…