Search Results for author: Shaojun Wang

Found 31 papers, 2 papers with code

PINGAN Omini-Sinitic at SemEval-2022 Task 4: Multi-prompt Training for Patronizing and Condescending Language Detection

no code implementations SemEval (NAACL) 2022 Ye Wang, Yanmeng Wang, Baishun Ling, Zexiang Liao, Shaojun Wang, Jing Xiao

This paper describes the second-placed system for subtask 2 and the ninth-placed system for subtask 1 in SemEval 2022 Task 4: Patronizing and Condescending Language Detection.

Binary Classification Classification +3

Sentinel: Attention Probing of Proxy Models for LLM Context Compression with an Understanding Perspective

1 code implementation29 May 2025 Yong Zhang, Yanwen Huang, Ning Cheng, Yang Guo, Yun Zhu, Yanmeng Wang, Shaojun Wang, Jing Xiao

Retrieval-augmented generation (RAG) enhances large language models (LLMs) with external context, but retrieved passages are often lengthy, noisy, or exceed input limits.

Decoder RAG +2

CHOrD: Generation of Collision-Free, House-Scale, and Organized Digital Twins for 3D Indoor Scenes with Controllable Floor Plans and Optimal Layouts

no code implementations15 Mar 2025 Chong Su, Yingbin Fu, Zheyuan Hu, Jing Yang, Param Hanji, Shaojun Wang, Xuan Zhao, Cengiz Öztireli, Fangcheng Zhong

We introduce CHOrD, a novel framework for scalable synthesis of 3D indoor scenes, designed to create house-scale, collision-free, and hierarchically structured indoor digital twins.

Indoor Scene Synthesis

Dynamic Attention-Guided Context Decoding for Mitigating Context Faithfulness Hallucinations in Large Language Models

no code implementations2 Jan 2025 Yanwen Huang, Yong Zhang, Ning Cheng, Zhitao Li, Shaojun Wang, Jing Xiao

Large language models (LLMs) often suffer from context faithfulness hallucinations, where outputs deviate from retrieved information due to insufficient context utilization and high output uncertainty.

Computational Efficiency

Rethinking Layer Removal: Preserving Critical Components with Task-Aware Singular Value Decomposition

no code implementations31 Dec 2024 Kainan Liu, Yong Zhang, Ning Cheng, Zhitao Li, Shaojun Wang, Jing Xiao

Layer removal has emerged as a promising approach for compressing large language models (LLMs) by leveraging redundancy within layers to reduce model size and accelerate inference.

Learning to Adapt to Low-Resource Paraphrase Generation

no code implementations22 Dec 2024 Zhigen Li, Yanmeng Wang, Rizhao Fan, Ye Wang, Jianfeng Li, Shaojun Wang

LAPA has three-stage training on three types of related resources to solve this problem: 1. pre-training PLMs on unsupervised corpora, 2. inserting an adapter layer and meta-training on source domain labeled data, and 3. fine-tuning adapters on a small amount of target domain labeled data.

Meta-Learning Paraphrase Generation

Enhancing Dual-Encoders with Question and Answer Cross-Embeddings for Answer Retrieval

no code implementations Findings (EMNLP) 2021 Yanmeng Wang, Jun Bai, Ye Wang, Jianfei Zhang, Wenge Rong, Zongcheng Ji, Shaojun Wang, Jing Xiao

To keep independent encoding of questions and answers during inference stage, variational auto-encoder is further introduced to reconstruct answers (questions) from question (answer) embeddings as an auxiliary task to enhance QA interaction in representation learning in training stage.

Question Answering Representation Learning +2

A Study of Different Ways to Use The Conformer Model For Spoken Language Understanding

no code implementations8 Apr 2022 Nick J. C. Wang, Shaojun Wang, Jing Xiao

In this paper, we compare different ways to combine ASR and NLU, in particular using a single Conformer model with different ways to use its components, to better understand the strengths and weaknesses of each approach.

Spoken Language Understanding

Adding Connectionist Temporal Summarization into Conformer to Improve Its Decoder Efficiency For Speech Recognition

no code implementations8 Apr 2022 Nick J. C. Wang, Zongfeng Quan, Shaojun Wang, Jing Xiao

The Conformer model is an excellent architecture for speech recognition modeling that effectively utilizes the hybrid losses of connectionist temporal classification (CTC) and attention to train model parameters.

Decoder speech-recognition +1

PINGAN Omini-Sinitic at SemEval-2021 Task 4:Reading Comprehension of Abstract Meaning

no code implementations SEMEVAL 2021 Ye Wang, Yanmeng Wang, Haijun Zhu, Bo Zeng, Zhenghong Hao, Shaojun Wang, Jing Xiao

This paper describes the winning system for subtask 2 and the second-placed system for subtask 1 in SemEval 2021 Task 4: ReadingComprehension of Abstract Meaning.

Denoising Language Modeling +2

Structure Controllable Text Generation

no code implementations1 Jan 2021 Liming Deng, Long Wang, Binzhu WANG, Jiang Qian, Bojin Zhuang, Shaojun Wang, Jing Xiao

Controlling the presented forms (or structures) of generated text are as important as controlling the generated contents during neural text generation.

Text Generation

KETG: A Knowledge Enhanced Text Generation Framework

no code implementations1 Jan 2021 Yan Cui, Xi Chen, Jiang Qian, Bojin Zhuang, Shaojun Wang, Jing Xiao

Embedding logical knowledge information into text generation is a challenging NLP task.

Diversity Text Generation

Contextualized Emotion Recognition in Conversation as Sequence Tagging

no code implementations1 Jul 2020 Yan Wang, Jiayu Zhang, Jun Ma, Shaojun Wang, Jing Xiao

Emotion recognition in conversation (ERC) is an important topic for developing empathetic machines in a variety of areas including social opinion mining, health-care and so on.

Emotion Classification Emotion Recognition in Conversation +1

BS-NAS: Broadening-and-Shrinking One-Shot NAS with Searchable Numbers of Channels

no code implementations22 Mar 2020 Zan Shen, Jiang Qian, Bojin Zhuang, Shaojun Wang, Jing Xiao

One-Shot methods have evolved into one of the most popular methods in Neural Architecture Search (NAS) due to weight sharing and single training of a supernet.

Neural Architecture Search

A simple discriminative training method for machine translation with large-scale features

no code implementations15 Sep 2019 Tian Xia, Shaodan Zhai, Shaojun Wang

Margin infused relaxed algorithms (MIRAs) dominate model tuning in statistical machine translation in the case of large scale features, but also they are famous for the complexity in implementation.

Machine Translation Translation

Plackett-Luce model for learning-to-rank task

no code implementations15 Sep 2019 Tian Xia, Shaodan Zhai, Shaojun Wang

List-wise based learning to rank methods are generally supposed to have better performance than point- and pair-wise based.

Learning-To-Rank model

Analysis of Regression Tree Fitting Algorithms in Learning to Rank

no code implementations12 Sep 2019 Tian Xia, Shaodan Zhai, Shaojun Wang

In learning to rank area, industry-level applications have been dominated by gradient boosting framework, which fits a tree using least square error principle.

Learning-To-Rank regression

Automatic Acrostic Couplet Generation with Three-Stage Neural Network Pipelines

no code implementations15 Jun 2019 Haoshen Fan, Jie Wang, Bojin Zhuang, Shaojun Wang, Jing Xiao

In this paper, we comprehensively study on automatic generation of acrostic couplet with the first characters defined by users.

Cultural Vocal Bursts Intensity Prediction Re-Ranking

A Syllable-Structured, Contextually-Based Conditionally Generation of Chinese Lyrics

no code implementations15 Jun 2019 Xu Lu, Jie Wang, Bojin Zhuang, Shaojun Wang, Jing Xiao

This paper presents a novel, syllable-structured Chinese lyrics generation model given a piece of original melody.

A Hierarchical Attention Based Seq2seq Model for Chinese Lyrics Generation

no code implementations15 Jun 2019 Haoshen Fan, Jie Wang, Bojin Zhuang, Shaojun Wang, Jing Xiao

In this paper, we comprehensively study on context-aware generation of Chinese song lyrics.

Sentence

Slim Embedding Layers for Recurrent Neural Language Models

no code implementations27 Nov 2017 Zhongliang Li, Raymond Kulhanek, Shaojun Wang, Yunxin Zhao, Shuang Wu

When the vocabulary size is large, the space taken to store the model parameters becomes the bottleneck for the use of recurrent neural language models.

Language Modeling Language Modelling

Une m\'ethode discriminant formation simple pour la traduction automatique avec Grands Caract\'eristiques

no code implementations JEPTALNRECITAL 2015 Tian Xia, Shaodan Zhai, Zhongliang Li, Shaojun Wang

Marge infus{\'e} algorithmes d{\'e}tendus (MIRAS) dominent mod{\`e}le de tuning dans la traduction automatique statistique dans le cas des grandes caract{\'e}ristiques de l{'}{\'e}chelle, mais ils sont {\'e}galement c{\'e}l{\`e}bres pour la complexit{\'e} de mise en {\oe}uvre.

Direct 0-1 Loss Minimization and Margin Maximization with Boosting

no code implementations NeurIPS 2013 Shaodan Zhai, Tian Xia, Ming Tan, Shaojun Wang

We propose a boosting method, DirectBoost, a greedy coordinate descent algorithm that builds an ensemble classifier of weak classifiers through directly minimizing empirical classification error over labeled training examples; once the training classification error is reduced to a local coordinatewise minimum, DirectBoost runs a greedy coordinate ascent algorithm that continuously adds weak classifiers to maximize any targeted arbitrarily defined margins until reaching a local coordinatewise maximum of the margins in a certain sense.

Classification General Classification

Cannot find the paper you are looking for? You can Submit a new open access paper.