Peng Xu

19

Papers

509

Total Citations

Papers (19)

OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models

Sub-sampled Newton Methods with Non-uniform Sampling

NeurIPS 2016arXiv

BESA: Pruning Large Language Models with Blockwise Parameter-Efficient Sparsity Allocation

LAKE-RED: Camouflaged Images Generation by Latent Background Knowledge Retrieval-Augmented Diffusion

AdsQA: Towards Advertisement Video Understanding

MMT-Bench: A Comprehensive Multimodal Benchmark for Evaluating Large Vision-Language Models Towards Multitask AGI

PIVOT: Iterative Visual Prompting Elicits Actionable Knowledge for VLMs

SketchMate: Deep Hashing for Million-Scale Human Sketch Retrieval

Trust Region Based Adversarial Attack on Neural Networks

DiffRate : Differentiable Compression Rate for Efficient Vision Transformers

DeepChange: A Long-Term Person Re-Identification Benchmark with Clothes Change

CVFusion: Cross-View Fusion of 4D Radar and Camera for 3D Object Detection

SCKD: Semi-Supervised Cross-Modality Knowledge Distillation for 4D Radar Object Detection

p-Laplacian Adaptation for Generative Pre-trained Vision-Language Models

Adaptive Multi-Modal Cross-Entropy Loss for Stereo Matching

InstructRetro: Instruction Tuning post Retrieval-Augmented Pretraining

GIANT: Globally Improved Approximate Newton Method for Distributed Optimization

Factuality Enhanced Language Models for Open-Ended Text Generation

Exploring the Limits of Domain-Adaptive Training for Detoxifying Large-Scale Language Models