Mingqing Zhang*, Haisong Gong*, Qiang Liu, Shu Wu, Liang Wang
arXiv, December, 2024
This paper introduces S2MAD, which investigate the use of multi-agent debate for rumor detection in social media within the context of breaking event scenarios.
Huanhuan Ma, Haisong Gong, Xiaoyuan Yi, Xing Xie, Dongkuan Xu
submitting, September, 2024
We propose the Core Sentiment Inventory (CSI), a novel tool inspired by the Implicit Association Test, to evaluate sentiment tendencies in Large Language Models.
Haisong Gong, Huanhuan Ma, Qiang Liu, Shu Wu, Liang Wang
arXiv, July, 2024
This paper introduces EACon where evidence abstraction and claim deconstruction are proposed to improve LLM’s performance in claim verification.
Lianwei Yang, Haisong Gong, Qingyi Gu
arXiv, July, 2024
This paper introduces Distribution-Friendly and Outlier-Aware Post-training Quantization method for Vision Transformers, named DopQ-ViT. It includes Tan Quatizer (TanQ) and Median as the Optimal Scaling Factor (MOSF).
Lianwei Yang, Zhikai Li, Junrui Xiao, Haisong Gong, Qingyi Gu
ICIP2024, June, 2024
This paper presents MGRA (Mixed Granularity Reconstruction Quantization) to improve post-training quantization for vision transformer.
Haisong Gong, Qiang Liu, Shu Wu, Liang Wang
AAAI2024, March, 2024
This paper presents TGM-DLM, the first diffusion language model designed for SMILE-based molecule generation, replacing traditional auto-regressive models. It highlights improved data efficiency, achieving superior results with a reduced amount of training data.
Recommended citation: Gong, H., Liu, Q., Wu, S., & Wang, L. (2024). Text-Guided Molecule Generation with Diffusion Language Model. Proceedings of the AAAI Conference on Artificial Intelligence, 38(1), 109-117.
Download Paper Code Repository
Haisong Gong, Weizhi Xu, Shu Wu, Qiang Liu,Liang Wang
AAAI2024, March, 2024
This paper presents HeterFC, where word-level heterogenous is build for graph reasoning. An attention-based method is utilized to integrate information. Multitask loss function is proposed to account for potential inaccuracies in evidence retrieval.
Recommended citation: Gong, H., Xu, W., Wu, S., Liu, Q., & Wang, L. (2024). Heterogeneous Graph Reasoning for Fact Checking over Texts and Tables. Proceedings of the AAAI Conference on Artificial Intelligence, 38(1), 100-108.
Download Paper Code Repository
Ronghao Dang, Haisong Gong, Yuxin Wang, Tianyi Huang, Zhuofan Shi, Xinwen Zhang, Yunfeng Wu, Yi Sun, Peng Qi
CCC, July, 2021
Develop an entire eel-like fish with a fully soft flexible body, which is composed of four fiber-reinforced, bidirectionally bending, fluidic elastomer actuators (FEAs) as its tail.