Chau Pham

selected publications

1. preprint

   Let Models Speak Ciphers: Multiagent Debate through Embeddings

   Chau Pham*, Boyi Liu*, Yingxiang Yang, Zhengyu Chen, Tianyi Liu, Jianbo Yuan, Bryan A. Plummer, Zhaoran Wang, and Hongxia Yang

   preprint

   Abstract PDF

   Discussion and debate among Large Language Models (LLMs) have gained considerable attention due to their potential to enhance the reasoning ability of LLMs. Although natural language is an obvious choice for communication due to LLM’s language understanding capability, the token sampling step needed when generating natural language poses a potential risk of information loss, as it uses only one token to represent the model’s belief across the entire vocabulary. In this paper, we introduce a communication regime named CIPHER (Communicative Inter-Model Protocol Through Embedding Representation) to address this issue. Specifically, we remove the token sampling step from LLMs and let them communicate their beliefs across the vocabulary through the expectation of the raw transformer output embeddings. Remarkably, by deviating from natural language, CIPHER offers an advantage of encoding a broader spectrum of information without any modification to the model weights. While the state-of-the-art LLM debate methods using natural language outperforms traditional inference by a margin of 1.5-8%, our experiment results show that CIPHER debate further extends this lead by 1-3.5% across five reasoning tasks and multiple open-source LLMs of varying sizes. This showcases the superiority and robustness of embeddings as an alternative "language" for communication among LLMs.

2. WACV

   MixtureGrowth: Growing Neural Networks by Recombining Learned Parameters

   Chau Pham*, Piotr Teterwak*, Soren Nelson*, and Bryan A. Plummer

   In IEEE Winter Conference on Applications of Computer Vision (WACV) 2024

   Abstract PDF

   Most deep neural networks are trained under fixed network architectures and require retraining when the architecture changes. If expanding the network’s size is needed, it is necessary to retrain from scratch, which is expensive. To avoid this, one can grow from a small network by adding random weights over time to gradually achieve the target network size. However, this naive approach falls short in practice as it brings too much noise to the growing process. Prior work tackled this issue by leveraging the already learned weights and training data for generating new weights through conducting a computationally expensive analysis step. In this paper, we introduce MixtureGrowth, a new approach to growing networks that circumvents the initialization overhead in prior work. Before growing, each layer in our model is generated with a linear combination of parameter templates. Newly grown layer weights are generated by using a new linear combination of existing templates for a layer. On one hand, these templates are already trained for the task, providing a strong initialization. On the other, the new coefficients provide flexibility for the added layer weights to learn something new. We show that our approach boosts top-1 accuracy over the state-of-the-art by 2-2.5% on CIFAR-100 and ImageNet datasets, while achieving comparable performance with fewer FLOPs to a larger network trained from scratch. Code is available at https://github.com/chaudatascience/mixturegrowth.

3. NeurIPS

   CHAMMI: A benchmark for channel-adaptive models in microscopy imaging

   Zitong Chen*, Chau Pham*, Siqi Wang, Michael Doron, Nikita Moshkov, Bryan A. Plummer, and Juan C. Caicedo

   In Advances in Neural Information Processing Systems (NeurIPS) Track on Datasets and Benchmarks 2023

   Abstract HTML PDF Code

   Most neural networks assume that input images have a fixed number of channels (three for RGB images). However, there are many settings where the number of channels may vary, such as microscopy images where the number of channels changes depending on instruments and experimental goals. Yet, there has not been a systemic attempt to create and evaluate neural networks that are invariant to the number and type of channels. As a result, trained models remain specific to individual studies and are hardly reusable for other microscopy settings. In this paper, we present a benchmark for investigating channel-adaptive models in microscopy imaging, which consists of 1) a dataset of varied-channel single-cell images, and 2) a biologically relevant evaluation framework. In addition, we adapted several existing techniques to create channel-adaptive models and compared their performance on this benchmark to fixed-channel, baseline models. We find that channel-adaptive models can generalize better to out-of-domain tasks and can be computationally efficient. We contribute a curated dataset and an evaluation API to facilitate objective comparisons in future research and applications.

4. IEEE Bigdata

   Graph Adversarial Attacks and Defense: An Empirical Study on Citation Graph

   Chau Pham, Vung Pham, and Tommy Dang

   In 2020 IEEE International Conference on Big Data (Big Data) 2020

   Abstract

   This paper details the methodologies and decisions making processes used while developing the attacking and defending models for the Graph Adversarial Attacks and Defense applied to a large citation graph. To handle the large graphs, our attack strategy is twofold: 1) randomly attack the structure first, 2) keep the structure unchanged, then continue the attack on the features using the gradient-based method. On the other hand, the defender is based on 1) filtering and normalizing the feature data, 2) applying the Graph Convolutional Network model, and 3) selecting the models with the highest accuracy and robustness based on our own attacking data. We applied these strategies in KDD Cup 2020 on Graph Adversarial Attacks and Defense dataset. The attacker can drop the accuracy of a surrogate 2-layer Graph Convolutional Network model from 60% to 30% on the test set. Our defending model has 68% accuracy on the validated data and has 89% of the target labels remained the same while adding fake nodes, generated by our attacking method, to the graph.

5. IEEE Bigdata

   Road Damage Detection and Classification with Detectron2 and Faster R-CNN

   Vung Pham, Chau Pham, and Tommy Dang

   In 2020 IEEE International Conference on Big Data (Big Data) 2020

   Abstract

   The road is vital for many aspects of life, and road maintenance is crucial for human safety. One of the critical tasks to allow timely repair of road damages is to quickly and efficiently detect and classify them. This work details the strategies and experiments evaluated for these tasks. Specifically, we evaluate Detectron2’s implementation of Faster R-CNN using different base models and configurations. We also experiment with these approaches using the Global Road Damage Detection Challenge 2020, A Track in the IEEE Big Data 2020 Big Data Cup Challenge dataset. The results show that the X101-FPN base model for Faster R-CNN with Detectron2’s default configurations is efficient and general enough to be transferable to different countries in this challenge. This approach results in F1 scores of 51.0% and 51.4% for the test1 and test2 sets of the challenge, respectively. Though the visualizations show good prediction results, the F1 scores are low. Therefore, we also evaluate the prediction results against the existing annotations and discover some discrepancies. Thus, we also suggest strategies to improve the labeling process for this dataset.