ExPLoRA: Parameter-Efficient Extended Pre-Training to Adapt Vision Transformers under Domain Shiftsby Samar Khanna, Medhanie Irgau,…
Updating CLIP to Prefer Descriptions Over Captionsby Amir Zur, Elisa Kreiss, Karel D'Oosterlinck, Christopher Potts,…
Parameter-Efficient Active Learning for Foundational modelsby Athmanarayanan Lakshmi Narayanan, Ranganath Krishnan, Amrutha Machireddy, Mahesh SubedarFirst…
Differentially Private Fine-Tuning of Diffusion Modelsby Yu-Lin Tsai, Yizhe Li, Zekai Chen, Po-Yu Chen, Chia-Mu…
SeamlessExpressiveLM: Speech Language Model for Expressive Speech-to-Speech Translation with Chain-of-Thoughtby Hongyu Gong, Bandhav VeluriFirst submitted…
Towards Cross-modal Backward-compatible Representation Learning for Vision-Language Modelsby Young Kyun Jang, Ser-nam LimFirst submitted to…
Lightweight Spatial Modeling for Combinatorial Information Extraction From Documentsby Yanfei Dong, Lambert Deng, Jiazheng Zhang,…
SPAFIT: Stratified Progressive Adaptation Fine-tuning for Pre-trained Large Language Modelsby Samir Arora, Liangliang WangFirst submitted…
HydraLoRA: An Asymmetric LoRA Architecture for Efficient Fine-Tuningby Chunlin Tian, Zhan Shi, Zhijiang Guo, Li…
Spatio-Temporal Side Tuning Pre-trained Foundation Models for Video-based Pedestrian Attribute Recognitionby Xiao Wang, Qian Zhu,…