Zongyu (Johnson) Lin「林宗裕」

I am a first-year CS Ph.D Student at UCLA, co-advised by Prof. Yizhou Sun and Prof. Kaiwei Chang. Before coming to UCLA, I have spent a year in Moonshot.AI (one of the earliest core members), working as a full-time research engineer working on LLM and VideoGen. I was one of the major contributor of training large language models with extremely long context, achieving the state-of-the-art performance on many long context tasks compared with GPT4 and Claude2 at that time, and also participated in visual generation project. I completed my bachelor's degree of Electronic Enginnering at Tsinghua University. Luckily, I have worked with Prof. Zhilin Yang, Prof. Yong Li, Prof. Hanan Samet and Prof. Cyrus Shahabi.

My research interest lies broadly in natural language processing and general machine learning. I have done some work including self-training, instruction finetuning and zero-shot task generalization of LLMs. Most Recently, I am interested in (1) studying the self-evolution paradigm of large language models as well as (2) exploring scalable architectures and recipes for multi-modal generation. Feel free to contact me for chat or discussion if you are also interested in these topics.

Email: lzyxx17 [at] gmail.com

Twitter   Google Scholar   Linkedin

profile photo
News
  • 2024-6 Our new paper on contradiction retrieval: SPARSECL: Sparse Contrastive Learning for Contradiction Retrieval is available now at preprint.
  • 2024-6 VideoPhy: Evaluating Physical Commonsense In Video Generation, get accepted by DMLR@ICML 2024, please check our preprint.
  • Experience

    Research Intern, Apple, 2024

    Research Engineer, Moonshot.AI 2023

    Quant Researcher Intern, Ubiquant, Top Hedge Fund in China. 2022

    Research Intern, Sensetime, China, 2021

    Research Topic

    My research interest lies broadly in natural language processing and general machine learning. Most Recently, I am interested in

    (1) exploring scalable architectures and recipes for multi-modal generation;

    (2) improving the self-evolution of LLMs / VLMs;

    (3) improve better alignment with the physical world for vision generative models and vision language models

    Recent Work
    SPARSECL: Sparse Contrastive Learning for Contradiction Retrieval
    Haike Xu*, Zongyu Lin*,Yizhou Sun, Kai-Wei Chang, Piotr Indyk
    *Equal Contribution
    arXiv, 2024
    preprint, website, code

    Contradiction retrieval refers to identifying and extracting documents that explicitly disagree with or refute the content of a query, which is important to many downstream applications like fact checking and data cleaning. To retrieve contradiction argument to the query from large document corpora, existing methods such as similarity search and crossencoder models exhibit significant limitations. The former struggles to capture the essence of contradiction due to its inherent nature of favoring similarity, while the latter suffers from computational inefficiency, especially when the size of corpora is large. To address these challenges, we introduce a novel approach: SPARSECL that leverages specially trained sentence embeddings designed to preserve subtle, contradictory nuances between sentences. Our method utilizes a combined metric of cosine similarity and a sparsity function to efficiently identify and retrieve documents that contradict a given query. This approach dramatically enhances the speed of contradiction detection by reducing the need for exhaustive document comparisons to simple vector calculations. We validate our model using the Arguana dataset, a benchmark dataset specifically geared towards contradiction retrieval, as well as synthetic contradictions generated from the MSMARCO and HotpotQA datasets using GPT-4. Our experiments demonstrate the efficacy of our approach not only in contradiction retrieval with more than 30% accuracy improvements on MSMARCO and HotpotQA across different model architectures but also in applications such as cleaning corrupted corpora to restore high-quality QA retrieval. This paper outlines a promising direction for improving the accuracy and efficiency of contradiction retrieval in large-scale text corpora.

    VideoPhy: Evaluating Physical Commonsense In Video Generation
    Hritik Bansal*, Zongyu Lin*, Jing Zhou, Tianyi Xie, Zeshun Zong, Michal Yarom, Yonatan Bitton, Chenfanfu Jiang, Yizhou Sun, Kai-Wei Chang, Aditya Grover
    *Equal Contribution
    arXiv, 2024, accepted by DMLR@ICML 2024
    preprint, website, code

    Recent advances in internet-scale video data pretraining have led to the development of text-to-video generative models that can create high-quality videos across a broad range of visual concepts and styles. Due to their ability to synthesize realistic motions and render complex objects, these generative models have the potential to become general-purpose simulators of the physical world. However, it is unclear how far we are from this goal with the existing text-to-video generative models. To this end, we present VideoPhy, a benchmark designed to assess whether the generated videos follow physical commonsense for real-world activities (e.g. marbles will roll down when placed on a slanted surface). Specifically, we curate a list of 688 captions that involve interactions between various material types in the physical world (e.g., solid-solid, solid-fluid, fluid-fluid). We then generate videos conditioned on these captions from diverse state-of-the-art text-to-video generative models, including open models (e.g., VideoCrafter2) and closed models (e.g., Lumiere from Google, Pika). Further, our human evaluation reveals that the existing models severely lack the ability to generate videos adhering to the given text prompts, while also lack physical commonsense. Specifically, the best performing model, Pika, generates videos that adhere to the caption and physical laws for only 19.7% of the instances. VideoPhy thus highlights that the video generative models are far from accurately simulating the physical world. Finally, we also supplement the dataset with an auto-evaluator, VideoCon-Physics, to assess semantic adherence and physical commonsense at scale.

    A Universal Discriminator for Zero-Shot Generalization
    Haike Xu, Zongyu Lin, Jing Zhou, Yanan Zheng, Zhilin Yang
    ACL Long Paper, 2023

    Generative modeling has been the dominant approach for large-scale pretraining and zero-shot generalization. In this work, we challenge this convention by showing that discriminative approaches perform substantially better than generative ones on a large number of NLP tasks. Technically, we train a single discriminator to predict whether a text sample comes from the true data distribution, similar to GANs. Since many NLP tasks can be formulated as selecting from a few options, we use this discriminator to predict the option with the highest probability. This simple formulation achieves state-of-the-art zero-shot results on the T0 benchmark, outperforming T0 by 16.0%, 7.8%, and 11.5% respectively on different scales. Meanwhile, our approach requires minimal prompting efforts, which largely improves robustness and is essential for real-world applications.

    NOT ALL TASKS ARE BORN EQUAL: UNDERSTANDING ZERO-SHOT GENERALIZATION
    Jing Zhou, Zongyu Lin, Yanan Zheng, Zhilin Yang
    ICLR Spotlight, 2023

    Recent work has achieved remarkable zero-shot performance with multi-task prompted pretraining, but little has been understood. For the first time, we show that training on a small number of key tasks beats using all the training tasks, while removing these key tasks substantially hurts performance. We also find that these key tasks are mostly question answering (QA) tasks. We design a shuffle experiment to further show that training on these QA tasks leads to better cross-task generalization in multi-task learning under various training/test task splits. These novel findings combined deepen our understanding about zero-generalization— training on certain tasks such as QA encodes general knowledge transferable to a wide range of tasks, which explains the improved zero-shot performance in recent progress. In addition, to automate this procedure, we devise a method to identify and upsample key training tasks without observing the test tasks based on cross validation. Empirically, our approach achieves improved results across various model scales and tasks.

    Learning to Detect Noisy Labels Using Model-Based Features
    Zhihao Wang*, Zongyu Lin*, Peiqi Liu, Guidong Zheng, Junjie Wen, Xianxin Chen, Yujun Chen, Zhilin Yang (* First Co-Authors)
    Findings of EMNLP, 2023

    Label noise is ubiquitous in various machine learning scenarios such as self-labeling with model predictions and erroneous data annotation. Many existing approaches are based on heuristics such as sample losses, which might not be flexible enough to achieve optimal solutions. Meta learning based methods address this issue by learning a data selection function, but can be hard to optimize. In light of these pros and cons, we propose SENT (Selection-Enhanced Noisy label Training) that does not rely on meta learning while having the flexibility of being data-driven. SENT transfers the noise distribution to a clean set and trains a model to distinguish noisy labels from clean ones using model-based features. Empirically, on a wide range of tasks including text classification and speech recognition, SENT improves performance over strong baselines under the settings of self-training and label corruption.

    Publications
    Learning to Detect Noisy Labels Using Model-Based Features
    Zongyu Lin*, Zhihao Wang* Peiqi Liu, Guidong Zheng, Junjie Wen, Xianxin Chen, Yujun Chen, Zhilin Yang (* First Co-Authors)
    Findings of EMNLP, 2022 (To Appear)
    project page / arXiv /

    Label noise is ubiquitous in various machine learning scenarios such as self-labeling with model predictions and erroneous data annotation. Many existing approaches are based on heuristics such as sample losses, which might not be flexible enough to achieve optimal solutions. Meta learning based methods address this issue by learning a data selection function, but can be hard to optimize. In light of these pros and cons, we propose SENT (Selection-Enhanced Noisy label Training) that does not rely on meta learning while having the flexibility of being data-driven. SENT transfers the noise distribution to a clean set and trains a model to distinguish noisy labels from clean ones using model-based features. Empirically, on a wide range of tasks including text classification and speech recognition, SENT improves performance over strong baselines under the settings of self-training and label corruption.

    Hagen: Homophily-aware graph convolutional recurrent network for crime forecasting
    Zongyu Lin*, Chenyu Wang*, Guozhen Zhang, Xiaochen Yang, Jiao Sun, Mingxuan Yue, Cyrus Shahabi (* First Co-Authors)
    Proceedings of the AAAI Conference on Artificial Intelligence, 2022
    paper

    We propose an end-to-end graph convolutional recurrent network called HAGEN with several novel designs for crime prediction. Specifically, our framework could jointly capture the crime correlation between regions and the temporal crime dynamics by combining an adaptive region graph learning module with the Diffusion Convolution Gated Recurrent Unit (DCGRU). Based on the homophily assumption of GNN (i.e., graph convolution works better where neighboring nodes share the same label), we propose a homophily-aware constraint to regularize the optimization of the region graph so that neighboring region nodes on the learned graph share similar crime patterns.

    Vehicle Trajectory Recovery on Road Network Based on Traffic Camera Video Data
    Zongyu Lin, Guozhen Zhang, Zhiqun He, Jie Feng, Wei Wu, Yong Li
    Proceedings of the 29th International Conference on Advances in Geographic Information Systems, 2021
    paper

    We propose a general system to recover vehicle trajectories at the level of the road intersection, where a novel iterative framework is developed to combine both vehicle clustering and trajectory recovery tasks.

    HealthWalks: Sensing Fine-grained Individual Health Condition via Mobility Data
    Zongyu Lin, Shiqing Lyu, Hancheng Cao, Yuqiong Wei, Pan Hui, Hanan Samet, Yong Li
    In ACM International Joint Conference on Pervasive and Ubiquitous Computing (UBICOMP), 2020
    paper

    We propose a DFA-based model which can generate interpretable features automatically from raw mobility data for fine-grained health sensing.

    SUME: Semantic-enhanced Urban Mobility Network Embedding for User Demographic Inference
    Fengli Xu*, Zongyu Lin*, Tong Xia, Diansheng Guo, Yong Li (* Equal Contributions)
    In ACM International Joint Conference on Pervasive and Ubiquitous Computing (UBICOMP), 2020
    paper

    We propose a semantic-enhanced urban mobility embedding model for user profiling, and reveal meaningful patterns in all spatial, temporal and urban structure domains.

    CrimeForecaster: Crime Prediction by Exploiting the Neighborhoods’ Spatiotemporal Dependencies
    Jiao Sun, Mingxuan Yue, Zongyu Lin, Xiaochen Yang, Gabe Kahn, Luciano Nocera, Cyrus Shahabi
    The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD) , 2020 (To appear)

    We introduce a new end-to-end spatiotemporal learning framework dubbed CrimeForecaster that: 1) represents the geographical extents of neighborhoods and their correlations in a graph; 2) uses graph convolution to predict crimes.

    Selected Awards

    Comprehensive Outstanding Scholarship(~10/280), Tsinghua University. 2020

    Excellent Technology Innovation Scholarship, Tsinghua University. 2020

    First Prize in Software Design Contest, Department of Electronic Enginnering, Tsinghua University. 2018

    Hobbies

    Sports! I really enjoy playing ballgames like football and tennis. I am a big fan of Lionel Messi, Rafael Nadal and Stephen Curry! Also, I love running, swimming and hiking.

    Updated at June.2024. Thanks Jon Barron for this concise and beautiful template.