Adam Accumulation to Reduce Memory Footprints of both Activations and Gradients for Large-scale DNN Training

Running out of GPU memory has become a main bottleneck for large-scale DNN training. How to reduce the memory footprint during training has received intensive research attention. We find that previous gradient accumulation reduces activation memory but fails to be compatible with gradient memory reduction due to a contradiction between preserving gradients and releasing gradients. To address this issue, we propose a novel optimizer accumulation method for Adam, named Adam Accumulation (AdamA), which enables reducing both activation and gradient memory. Specifically, AdamA directly integrates gradients into optimizer states and accumulates optimizer states over micro-batches, so that gradients can be released immediately after use. We mathematically and experimentally demonstrate AdamA yields the same convergence properties as Adam. Evaluated on transformer-based models, AdamA achieves up to 23% memory reduction compared to gradient accumulation with less than 2% degradation in training throughput. Notably, AdamA can work together with memory reduction methods for optimizer states to fit 1.26x~3.14x larger models over PyTorch and DeepSpeed baseline on GPUs with different memory capacities.

Reversible Deep Neural Network Watermarking:Matching the Floating-point Weights

Static deep neural network (DNN) watermarking embeds watermarks into the weights of DNN model by irreversible methods, but this will cause permanent damage to watermarked model and can not meet the requirements of integrity authentication. For these reasons, reversible data hiding (RDH) seems more attractive for the copyright protection of DNNs. This paper proposes a novel RDH-based static DNN watermarking method by improving the non-reversible quantization index modulation (QIM). Targeting the floating-point weights of DNNs, the idea of our RDH method is to add a scaled quantization error back to the cover object. Two schemes are designed to realize the integrity protection and legitimate authentication of DNNs. Simulation results on training loss and classification accuracy justify the superior feasibility, effectiveness and adaptability of the proposed method over histogram shifting (HS).

Cross-supervised Dual Classifiers for Semi-supervised Medical Image Segmentation

Semi-supervised medical image segmentation offers a promising solution for large-scale medical image analysis by significantly reducing the annotation burden while achieving comparable performance. Employing this method exhibits a high degree of potential for optimizing the segmentation process and increasing its feasibility in clinical settings during translational investigations. Recently, cross-supervised training based on different co-training sub-networks has become a standard paradigm for this task. Still, the critical issues of sub-network disagreement and label-noise suppression require further attention and progress in cross-supervised training. This paper proposes a cross-supervised learning framework based on dual classifiers (DC-Net), including an evidential classifier and a vanilla classifier. The two classifiers exhibit complementary characteristics, enabling them to handle disagreement effectively and generate more robust and accurate pseudo-labels for unlabeled data. We also incorporate the uncertainty estimation from the evidential classifier into cross-supervised training to alleviate the negative effect of the error supervision signal. The extensive experiments on LA and Pancreas-CT dataset illustrate that DC-Net outperforms other state-of-the-art methods for semi-supervised segmentation. The code will be released soon.

Self-aware and Cross-sample Prototypical Learning for Semi-supervised Medical Image Segmentation

Consistency learning plays a crucial role in semi-supervised medical image segmentation as it enables the effective utilization of limited annotated data while leveraging the abundance of unannotated data. The effectiveness and efficiency of consistency learning are challenged by prediction diversity and training stability, which are often overlooked by existing studies. Meanwhile, the limited quantity of labeled data for training often proves inadequate for formulating intra-class compactness and inter-class discrepancy of pseudo labels. To address these issues, we propose a self-aware and cross-sample prototypical learning method (SCP-Net) to enhance the diversity of prediction in consistency learning by utilizing a broader range of semantic information derived from multiple inputs. Furthermore, we introduce a self-aware consistency learning method that exploits unlabeled data to improve the compactness of pseudo labels within each class. Moreover, a dual loss re-weighting method is integrated into the cross-sample prototypical consistency learning method to improve the reliability and stability of our model. Extensive experiments on ACDC dataset and PROMISE12 dataset validate that SCP-Net outperforms other state-of-the-art semi-supervised segmentation methods and achieves significant performance gains compared to the limited supervised training. Our code will come soon.

OR-NeRF: Object Removing from 3D Scenes Guided by Multiview Segmentation with Neural Radiance Fields

The emergence of Neural Radiance Fields (NeRF) for novel view synthesis has led to increased interest in 3D scene editing. One important task in editing is removing objects from a scene while ensuring visual reasonability and multiview consistency. However, current methods face challenges such as time-consuming object labelling, limited capability to remove specific targets, and compromised rendering quality after removal. This paper proposes a novel object-removing pipeline, named OR-NeRF, that can remove objects from 3D scenes with either point or text prompts on a single view, achieving better performance in less time than previous works. Our method uses a points projection strategy to rapidly spread user annotations to all views, significantly reducing the processing burden. This algorithm allows us to leverage the recent 2D segmentation model Segment-Anything (SAM) to predict masks with improved precision and efficiency. Additionally, we obtain colour and depth priors through 2D inpainting methods. Finally, our algorithm employs depth supervision and perceptual loss for scene reconstruction to maintain consistency in geometry and appearance after object removal. Experimental results demonstrate that our method achieves better editing quality with less time than previous works, considering both quality and quantity.

Interpretation of Time-Series Deep Models: A Survey

Deep learning models developed for time-series associated tasks have become more widely researched nowadays. However, due to the unintuitive nature of time-series data, the interpretability problem -- where we understand what is under the hood of these models -- becomes crucial. The advancement of similar studies in computer vision has given rise to many post-hoc methods, which can also shed light on how to explain time-series models. In this paper, we present a wide range of post-hoc interpretation methods for time-series models based on backpropagation, perturbation, and approximation. We also want to bring focus onto inherently interpretable models, a novel category of interpretation where human-understandable information is designed within the models. Furthermore, we introduce some common evaluation metrics used for the explanations, and propose several directions of future researches on the time-series interpretability problem. As a highlight, our work summarizes not only the well-established interpretation methods, but also a handful of fairly recent and under-developed techniques, which we hope to capture their essence and spark future endeavours to innovate and improvise.

Graph Meets LLM: A Novel Approach to Collaborative Filtering for Robust Conversational Understanding

Conversational AI systems (e.g. Alexa, Siri, Google Assistant, etc.) need to understand queries with defects to ensure robust conversational understanding and reduce user frictions. The defective queries are often induced by user ambiguities and mistakes, or errors in the automatic speech recognition (ASR) and natural language understanding (NLU). Personalized query rewriting (personalized QR) targets reducing defects in the torso and tail user query traffic, and it typically relies on an index of past successful user interactions with the conversational AI. This paper presents our "Collaborative Query Rewriting" approach that focuses on rewriting novel user interactions unseen in the user history. This approach builds a "user Feedback Interaction Graph" (FIG) consisting of historical user-entity interactions, and leverages multi-hop customer affinity to enrich each user's index (i.e. the Collaborative User Index) that would help cover future unseen defective queries. To counteract the precision degradation from the enlarged index, we introduced additional transformer layers to the L1 retrieval model and added multi-hop affinity and guardrail features to the L2 re-ranking model. Given the production constraints of storage cost and runtime retrieval latency, managing the size of the Collaborative User Index is important. As the user index can be pre-computed, we explored using a Large Language Model (LLM) for multi-hop customer affinity retrieval on the Video/Music domains. In particular, this paper looked into the Dolly-V2 7B model. Given limited user index size, We found the user index derived from fine-tuned Dolly-V2 generation significantly enhanced coverage of unseen user interactions. Consequently, this boosted QR performance on unseen user interactions compared to the graph traversal based user index.

DEGREE: Decomposition Based Explanation For Graph Neural Networks

Graph Neural Networks (GNNs) are gaining extensive attention for their application in graph data. However, the black-box nature of GNNs prevents users from understanding and trusting the models, thus hampering their applicability. Whereas explaining GNNs remains a challenge, most existing methods fall into approximation based and perturbation based approaches with suffer from faithfulness problems and unnatural artifacts, respectively. To tackle these problems, we propose DEGREE \degree to provide a faithful explanation for GNN predictions. By decomposing the information generation and aggregation mechanism of GNNs, DEGREE allows tracking the contributions of specific components of the input graph to the final prediction. Based on this, we further design a subgraph level interpretation algorithm to reveal complex interactions between graph nodes that are overlooked by previous methods. The efficiency of our algorithm can be further improved by utilizing GNN characteristics. Finally, we conduct quantitative and qualitative experiments on synthetic and real-world datasets to demonstrate the effectiveness of DEGREE on node classification and graph classification tasks.

Integer or Floating Point? New Outlooks for Low-Bit Quantization on Large Language Models

Efficient deployment of large language models (LLMs) necessitates low-bit quantization to minimize model size and inference cost. While low-bit integer formats (e.g., INT8/INT4) have been the conventional choice, emerging low-bit floating-point formats (e.g., FP8/FP4) offer a compelling alternative and are gaining support from cutting-edge hardware, such as NVIDIA's H100 GPU. However, the superiority of low-bit INT versus FP formats for quantization on LLMs remains unclear. In this study, we conduct a comparative analysis of INT and FP quantization with the same bit-width, revealing that the optimal quantization format varies across different layers due to the complexity and diversity of tensor distribution. Consequently, we advocate the Mixture of Formats Quantization (MoFQ), which selects the optimal format on a layer-wise basis. This simple yet effective approach achieves state-of-the-art results in both weight-only (W-only) and weight-activation (WA) post-training quantization scenarios when tested on LLaMA across various tasks. In 4-bit W-only quantization, MoFQ surpasses GPTQ without complex hyperparameter tuning and with an order of magnitude faster quantization speed. While in 8-bit WA quantization, MoFQ significantly outperforms INT/FP-only methods, achieving performance close to the full precision model. Notably, MoFQ incurs no hardware overhead compared to INT/FP-only quantization, as the bit-width remains unchanged.

The Ways of Words: The Impact of Word Choice on Information Engagement and Decision Making

Little research has explored how information engagement (IE), the degree to which individuals interact with and use information in a manner that manifests cognitively, behaviorally, and affectively. This study explored the impact of phrasing, specifically word choice, on IE and decision making. Synthesizing two theoretical models, User Engagement Theory UET and Information Behavior Theory IBT, a theoretical framework illustrating the impact of and relationships among the three IE dimensions of perception, participation, and perseverance was developed and hypotheses generated. The framework was empirically validated in a large-scale user study measuring how word choice impacts the dimensions of IE. The findings provide evidence that IE differs from other forms of engagement in that it is driven and fostered by the expression of the information itself, regardless of the information system used to view, interact with, and use the information. The findings suggest that phrasing can have a significant effect on the interpretation of and interaction with digital information, indicating the importance of expression of information, in particular word choice, on decision making and IE. The research contributes to the literature by identifying methods for assessment and improvement of IE and decision making with digital text.