Proposed a salience-aware learning method for table-based fact verification, achieving SOTA on TabFact.

Fei Wang, Kexuan Sun, Jay Pujara et al.

EMA: ensemble-based membership inference for robust data removal auditing, outperforming KS-distance methods.

Yangsibo Huang, Xiaoxiao Li, Kai Li

Utilizing deep learning to address the accuracy and scalability issues in musical version identification.

Furkan Yesiler, Guillaume Doras, Rachel M. Bittner et al.

This paper introduces instruction tuning on a 137B parameter model, significantly improving zero-shot performance across 60 NLP tasks, outperforming GPT-3 on many benchmarks.

Jason Wei, Maarten Bosma, Vincent Y. Zhao et al.

Introduces MULTI-EURLEX, a multilingual, multi-label legal dataset, and explores zero-shot cross-lingual transfer using models like XLM-ROBERTA and MT5 with adaptation strategies.

Ilias Chalkidis, Manos Fergadiotis, Ion Androutsopoulos

Introduces CO3D dataset and NerFormer, advancing real-world 3D reconstruction and novel view synthesis.

Jeremy Reizenstein, Roman Shapovalov, Philipp Henzler et al.

Introduces FinQA, a financial QA dataset with expert-annotated multi-step numerical reasoning, highlighting model performance gaps.

Zhiyu Chen, Wenhu Chen, Charese Smiley et al.

Proposed a NeRF-based self-calibration algorithm, significantly improving camera parameter estimation and rendering quality.

Yoonwoo Jeong, Seokjun Ahn, Christopher Choy et al.

Improving query representations in dense retrieval using pseudo relevance feedback, significantly enhancing accuracy.

HongChien Yu, Chenyan Xiong, Jamie Callan

Integrates Wasserstein GAN with PINNs for PDE uncertainty quantification, using groupsort activations to enhance discriminator capacity.

Yihang Gao, Michael K. Ng

ALiBi introduces linear distance biases in attention scores, enabling models trained on short sequences to extrapolate to longer inputs efficiently.

Ofir Press, Noah A. Smith, Mike Lewis

Proposes a cross-modal contrastive learning framework for video domain adaptation, leveraging RGB and optical flow features, achieving state-of-the-art results on UCF-HMDB and EPIC-Kitchens.

Donghyun Kim, Yi-Hsuan Tsai, Bingbing Zhuang et al.

DROID-SLAM integrates deep learning with classical optimization, using Dense Bundle Adjustment for high-precision, robust monocular, stereo, and RGB-D visual SLAM.

Zachary Teed, Jia Deng

Integrates formal verification with adversarial training to enhance DNN robustness, achieving 20% accuracy improvement against attacks on CIFAR-10.

Wenjie Ruan, Xinping Yi, Xiaowei Huang

DenseTNT is an end-to-end, anchor-free trajectory prediction model using dense goal sets, achieving top performance on Argoverse and Waymo datasets.

Junru Gu, Chen Sun, Hang Zhao

This paper explores sentence embeddings from T5, proposing three extraction methods, outperforming Sentence-BERT and SimCSE, with scaling improving performance.

Jianmo Ni, Gustavo Hernández Ábrego, Noah Constant et al.

PoinTr employs geometry-aware Transformers for point cloud completion, outperforming SOTA with significant margins.

Xumin Yu, Yongming Rao, Ziyi Wang et al.

StructDepth leverages indoor structural regularities (Manhattan model and co-planar constraints) to improve monocular depth estimation.

Boying Li, Yuan Huang, Zeyu Liu et al.

Proposes neural operators combining integral kernels and nonlinear activations for PDE operator approximation, achieving universal approximation and discretization invariance.

Nikola Kovachki, Zongyi Li, Burigede Liu et al.

SAMP uses cVAE and A* algorithm for scene-aware motion prediction, generating diverse action styles.

Mohamed Hassan, Duygu Ceylan, Ruben Villegas et al.