The Research Desk.

The most upvoted and starred AI research crossing the community today.

Last Brew Time: Jul 22, 2026, 10:54 AM PT

X.com Research Buzz

NLP
Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMs
X.com
11416

Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMs

Joseba Fernandez de Landa, Carla Perez-Almendros, Jose Camacho-Collados

HiTZ Center - Ixa, University of the Basque Country EHU, Cardiff University

AlphaXiv Trending

NLP
AlphaXiv
234

Language model harnesses are compositional generalizers

Alex Zhang, Omar Khattab

AlphaXiv
55

Introducing Laguna S 2.1

Poolside

NLP
LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks
AlphaXiv
47

LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks

Tianzhu Ye, Li Dong, Guanheng Chen

Microsoft Research, Tsinghua University, Peking University

Computer Vision
Exploration Matters for Escaping the Blur Trap in 3D Gaussian Splatting
AlphaXiv
30

Exploration Matters for Escaping the Blur Trap in 3D Gaussian Splatting

Chengbo Wang, Guozheng Ma, Jinhong Wu

Hunan University, Nanyang Technological University

Computer Vision
Masked Visual Actions for Unified World Modeling
AlphaXiv
30

Masked Visual Actions for Unified World Modeling

Hadi Alzayer, Wenlong Huang, Haonan Chen

Stanford University, University of Maryland, College Park, Harvard University

HuggingFace Daily Papers

Reinforcement Learning
Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing
HuggingFace
4

Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing

Laurin Wagner, Mario Zusag, Bernhard Thallinger

nyra labs

Machine Learning
Where Should Optimizer State Live? Tiered State Allocation for Memory-Efficient Mixture-of-Experts Training
HuggingFace
4

Where Should Optimizer State Live? Tiered State Allocation for Memory-Efficient Mixture-of-Experts Training

Nuemaan Malik

NLP
Subliminal Clocks: Latent Time Modelling in Diffusion Language Models
HuggingFace
3

Subliminal Clocks: Latent Time Modelling in Diffusion Language Models

Maximo Eduardo Rulli, Thomas Vaitses Fontanari, Simone Petruzzi, Federico Alvetreti, Giorgio Strano

Sapienza University of Rome

Delineate Anything v2: A Global Foundation Model for Field Delineation
HuggingFace
3

Delineate Anything v2: A Global Foundation Model for Field Delineation

Mykola Lavreniuk, Nataliia Kussul, Andrii Shelestov, Yevhenii Salii, Volodymyr Kuzin

NLP
EduPanel: A Three-Agent LLM Judge for Teaching Videos -- Reliability, Complementarity, and Human Trust Calibration
HuggingFace
3

EduPanel: A Three-Agent LLM Judge for Teaching Videos -- Reliability, Complementarity, and Human Trust Calibration

Jia-Kai Dong, Yi-Cheng Lin, Hung-yi Lee

台灣大學

NLP
Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges
HuggingFace
1

Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges

Tuo Liang, Zhe Hu, Disheng Liu, Jing Li, Yu Yin