Proposed a Bayesian causal inference method to address interference, treatment nonadherence, and missing outcomes.

Yuki Ohnishi, Arman Sabbaghi

Proposed DIFFNet uses HRNet with internal feature fusion and attention, outperforming SOTA on KITTI for monocular depth estimation.

Hang Zhou, David Greenwood, Sarah Taylor

StyleNeRF integrates NeRF with style modulation for high-res, multi-view consistent 3D-aware image synthesis, enabling interactive rendering and explicit control.

Jiatao Gu, Lingjie Liu, Peng Wang et al.

Proposed a Transformer-based model for fast, high-fidelity visually guided sound generation.

Vladimir Iashin, Esa Rahtu

This paper analyzes adversarial attacks on Gaussian process bandits, proposing white-box and black-box methods that successfully steer algorithms toward target regions with low budgets.

Eric Han, Jonathan Scarlett

Multitask prompted training (T0) achieves state-of-the-art zero-shot performance, outperforming models up to 16× larger on diverse NLP tasks.

Victor Sanh, Albert Webson, Colin Raffel et al.

Introduces BBQ, a handcrafted bias benchmark for question answering, revealing models' reliance on stereotypes especially under low-information conditions, with bias scores up to 100%.

Alicia Parrish, Angelica Chen, Nikita Nangia et al.

Proposes LSPE architecture with learnable structural and positional encodings, boosting molecular dataset performance up to 64.14%.

Vijay Prakash Dwivedi, Anh Tuan Luu, Thomas Laurent et al.

NeRS employs neural surface parameterization and BRDF modeling to achieve high-quality sparse-view 3D reconstruction with guaranteed watertightness and view-dependent reflectance.

Jason Y. Zhang, Gengshan Yang, Shubham Tulsiani et al.

P-Tuning v2 employs deep layered prompts, matching fine-tuning performance with only 0.1%-3% trainable parameters across various model scales and tasks.

Xiao Liu, Kaixuan Ji, Yicheng Fu et al.

Ego4D introduces a large-scale egocentric dataset and multi-task models, advancing first-person visual understanding.

Kristen Grauman, Andrew Westbury, Eugene Byrne et al.

Revises near-field distance limits for large arrays and RIS, introducing N·dF and 2D√N metrics, surpassing traditional Fraunhofer bounds.

Emil Björnson, Özlem Tugfe Demir, Luca Sanguinetti

Proposes RemBERT, a distilled multilingual evaluation metric reaching 92.6% of the large model's performance with only one-third parameters.

Amy Pu, Hyung Won Chung, Ankur P. Parikh et al.

Proposed Crystal Diffusion Variational Autoencoder (CDVAE) leverages diffusion processes and SE(3) equivariant GNNs to generate stable periodic materials, outperforming prior methods.

Tian Xie, Xiang Fu, Octavian-Eugen Ganea et al.

Proposes successor representation (SR)-based framework for automatic option discovery and composition, enhancing exploration and planning in reinforcement learning.

Marlos C. Machado, Andre Barreto, Doina Precup et al.

DuoRec employs contrastive regularization to address representation degeneration in sequential recommendation, improving embedding uniformity.

Ruihong Qiu, Zi Huang, Hongzhi Yin et al.

Carefully tuned recurrent off-policy RL matches or exceeds specialized POMDP algorithms across 18/21 environments, demonstrating strong baseline performance.

Tianwei Ni, Benjamin Eysenbach, Ruslan Salakhutdinov

ATISS uses an autoregressive Transformer to generate room layouts on 3D-FRONT, with up to 8× faster inference and better realism.

Despoina Paschalidou, Amlan Kar, Maria Shugrina et al.

Proposes a hybrid direct-iterative method using Schur complements and GPU-friendly Cholesky for large-scale KKT systems, outperforming LDLT.

Shaked Regev, Nai-Yuan Chiang, Eric Darve et al.

Proposes CKL framework with INVARIANTLAMA, UPDATEDLAMA, NEWLAMA datasets and FUAR metric; uses parameter expansion methods to improve knowledge retention and acquisition.

Joel Jang, Seonghyeon Ye, Sohee Yang et al.