MiROCL method excels on industrial and synthetic data, improving performance by over 30%.

Haozhe Wang, Chao Du, Panyan Fang et al.

All-optical neural networks can drastically reduce inference energy, achieving ~10^4× lower energy than electronic systems in large models, thanks to strong nonlinear materials and static weights.

Michał Matuszewski, Adam Prystupiuk, Andrzej Opala

Introduces Aria Digital Twin (ADT), a dataset with 200 egocentric sequences, supporting 3D detection, tracking, and scene understanding tasks.

Xiaqing Pan, Nicholas Charron, Yongqian Yang et al.

Proposed C3D mechanism uses data corruption and minimax estimators to ensure incentive-compatible collaborative mean estimation, achieving social penalty within twice the optimal.

Yiding Chen, Xiaojin Zhu, Kirthevasan Kandasamy

A multi-modal foundation model system enables zero-shot stylized 3D asset generation from abstract scene descriptions, outperforming baselines in semantic fidelity.

Ian Huang, Vrishab Krishna, Omoruyi Atekha et al.

Prodigy estimates D to optimize learning rates adaptively, outperforming D-Adaptation and nearing hand-tuned Adam in experiments.

Konstantin Mishchenko, Aaron Defazio

Mind2Web is a large-scale dataset with 137 real websites and 2000+ tasks, enabling web task automation via combined small and large language models.

Xiang Deng, Yu Gu, Boyuan Zheng et al.

AdaBatAL method improves active learning efficiency with adaptive batch sizes, significantly enhancing Bayesian optimization performance.

Masaki Adachi, Satoshi Hayakawa, Martin Jørgensen et al.

REMEMBERER framework enhances LLM's RL capabilities with long-term experience memory, achieving a 4% success rate increase.

Danyang Zhang, Lu Chen, Situo Zhang et al.

MusicGen employs a single Transformer with multi-stream codebook interleaving for high-quality text- and melody-conditioned music synthesis.

Jade Copet, Felix Kreuk, Itai Gat et al.

Proposes a hypernetwork-based controllable multi-objective re-ranking (CMR) framework enabling real-time preference adjustment without retraining, improving recommendation quality.

Sirui Chen, Yuan Wang, Zijing Wen et al.

PandaLM is an automatic evaluation benchmark for LLM instruction tuning, achieving 93.75% GPT-3.5 evaluation capability with multi-dimensional subjective metrics.

Yidong Wang, Zhuohao Yu, Zhengran Zeng et al.

Large-scale human perception study reveals poor correlation of existing metrics with perceived realism; introduces DINOv2-ViT-L/14 for improved evaluation of diffusion models.

George Stein, Jesse C. Cresswell, Rasa Hosseinzadeh et al.

This paper proves transformers can implement broad machine learning algorithms and adaptively select algorithms across tasks, with near-optimal predictive performance.

Yu Bai, Fan Chen, Huan Wang et al.

RotorPy is a Python-based multirotor simulator with aerodynamics, enabling detailed UAV control and wind estimation studies.

Spencer Folk, James Paulos, Vijay Kumar

M$^3$IT dataset optimizes vision-language models with 40 datasets and 80 languages.

Lei Li, Yuwei Yin, Shicheng Li et al.

Proposed an improved hierarchical OPF algorithm for three-phase unbalanced networks, enhancing voltage safety, validated on IEEE 123-bus test.

Heng Liang, Xinyang Zhou, Changhong Zhao

This paper introduces the Randomly Pivoted Cholesky (RPCHOLESKY) kernel quadrature method, achieving near-optimal error rates with significantly improved computational efficiency.

Ethan N. Epperly, Elvira Moreno

DIFT leverages pre-trained diffusion models' implicit features for unsupervised semantic, geometric, and temporal correspondence, outperforming weakly-supervised methods.

Luming Tang, Menglin Jia, Qianqian Wang et al.

RAM model achieves high-accuracy zero-shot image tagging using large-scale image-text pair training.

Youcai Zhang, Xinyu Huang, Jinyu Ma et al.