Entity matching (EM) is the problem of determining whether two records refer to same real-world entity, which is crucial in data integration, e.g., for product catalogs or address databases. A major drawback of many EM approaches is their dependence on labelled examples. We thus focus on the challenging setting of zero-shot entity matching where no labelled examples are available for an unseen target dataset. Recently, large language models (LLMs) have shown promising results for zero-shot EM, but their low throughput and high deployment cost limit their applicability and scalability. We revisit the zero-shot EM problem with AnyMatch, a small language model fine-tuned in a transfer learning setup. We propose several novel data selection techniques to generate fine-tuning data for our model, e.g., by selecting difficult pairs to match via an AutoML filter, by generating additional attribute-level examples, and by controlling label imbalance in the data. We conduct an extensive evaluation of the prediction quality and deployment cost of our model, in a comparison to thirteen baselines on nine benchmark datasets. We find that AnyMatch provides competitive prediction quality despite its small parameter size: it achieves the second-highest F1 score overall, and outperforms several other approaches that employ models with hundreds of billions of parameters. Furthermore, our approach exhibits major cost benefits: the average prediction quality of AnyMatch is within 4.4% of the state-of-the-art method MatchGPT with the proprietary trillion-parameter model GPT-4, yet AnyMatch requires four orders of magnitude less parameters and incurs a 3,899 times lower inference cost (in dollars per 1,000 tokens).
Cross-domain recommendation (CDR) aims to improve recommendation accuracy in sparse domains by transferring knowledge from data-rich domains. However, existing CDR approaches often assume that user-item interaction data across domains is publicly available, neglecting user privacy concerns. Additionally, they experience performance degradation with sparse overlapping users due to their reliance on a large number of fully shared users for knowledge transfer. To address these challenges, we propose a Federated Prototype-based Contrastive Learning (CL) framework for Privacy Preserving CDR, called FedPCL-CDR. This approach utilizes non-overlapping user information and differential prototypes to improve model performance within a federated learning framework. FedPCL-CDR comprises two key modules: local domain (client) learning and global server aggregation. In the local domain, FedPCL-CDR first clusters all user data and utilizes local differential privacy (LDP) to learn differential prototypes, effectively utilizing non-overlapping user information and protecting user privacy. It then conducts knowledge transfer by employing both local and global prototypes returned from the server in a CL manner. Meanwhile, the global server aggregates differential prototypes sent from local domains to learn both local and global prototypes. Extensive experiments on four CDR tasks across Amazon and Douban datasets demonstrate that FedPCL-CDR surpasses SOTA baselines. We release our code at https://github.com/Lili1013/FedPCL CDR
Sequential recommendation aims to predict the next item a user is likely to prefer based on their sequential interaction history. Recently, text-based sequential recommendation has emerged as a promising paradigm that uses pre-trained language models to exploit textual item features to enhance performance and facilitate knowledge transfer to unseen datasets. However, existing text-based recommender models still struggle with two key challenges: (i) representing users and items with multiple attributes, and (ii) matching items with complex user interests. To address these challenges, we propose a novel model, Matching Attribute-aware Representations for Text-based Sequential Recommendation (MARS). MARS extracts detailed user and item representations through attribute-aware text encoding, capturing diverse user intents with multiple attribute-aware representations. It then computes user-item scores via attribute-wise interaction matching, effectively capturing attribute-level user preferences. Our extensive experiments demonstrate that MARS significantly outperforms existing sequential models, achieving improvements of up to 24.43% and 29.26% in Recall@10 and NDCG@10 across five benchmark datasets. Code is available at https://github.com/junieberry/MARS
In the realm of Adversarial Distillation (AD), strategic and precise knowledge transfer from an adversarially robust teacher model to a less robust student model is paramount. Our Dynamic Guidance Adversarial Distillation (DGAD) framework directly tackles the challenge of differential sample importance, with a keen focus on rectifying the teacher model's misclassifications. DGAD employs Misclassification-Aware Partitioning (MAP) to dynamically tailor the distillation focus, optimizing the learning process by steering towards the most reliable teacher predictions. Additionally, our Error-corrective Label Swapping (ELS) corrects misclassifications of the teacher on both clean and adversarially perturbed inputs, refining the quality of knowledge transfer. Further, Predictive Consistency Regularization (PCR) guarantees consistent performance of the student model across both clean and adversarial inputs, significantly enhancing its overall robustness. By integrating these methodologies, DGAD significantly improves upon the accuracy of clean data and fortifies the model's defenses against sophisticated adversarial threats. Our experimental validation on CIFAR10, CIFAR100, and Tiny ImageNet datasets, employing various model architectures, demonstrates the efficacy of DGAD, establishing it as a promising approach for enhancing both the robustness and accuracy of student models in adversarial settings.
Adapting pre-trained models to new tasks can exhibit varying effectiveness across datasets. Visual prompting, a state-of-the-art parameter-efficient transfer learning method, can significantly improve the performance of out-of-distribution tasks. On the other hand, linear probing, a standard transfer learning method, can sometimes become the best approach. We propose a log-likelihood ratio (LLR) approach to analyze the comparative benefits of visual prompting and linear probing. By employing the LLR score alongside resource-efficient visual prompts approximations, our cost-effective measure attains up to a 100-fold reduction in run time compared to full training, while achieving prediction accuracies up to 91%. The source code is available at https://github.com/IBM/VP-LLR.
Pre-trained large vision-language models (VLMs) like CLIP have revolutionized visual representation learning using natural language as supervisions, and demonstrated promising generalization ability. In this work, we propose ViP, a novel visual symptom-guided prompt learning framework for medical image analysis, which facilitates general knowledge transfer from CLIP. ViP consists of two key components: a visual symptom generator (VSG) and a dual-prompt network. Specifically, VSG aims to extract explicable visual symptoms from pre-trained large language models, while the dual-prompt network utilizes these visual symptoms to guide the training on two learnable prompt modules, i.e., context prompt and merge prompt, which effectively adapts our framework to medical image analysis via large VLMs. Extensive experimental results demonstrate that ViP can outperform state-of-the-art methods on two challenging datasets.
Multi-task learning (MTL) is a machine learning paradigm that aims to improve the generalization performance of a model on multiple related tasks by training it simultaneously on those tasks. Unlike MTL, where the model has instant access to the training data of all tasks, continual learning (CL) involves adapting to new sequentially arriving tasks over time without forgetting the previously acquired knowledge. Despite the wide practical adoption of CL and MTL and extensive literature on both areas, there remains a gap in the theoretical understanding of these methods when used with overparameterized models such as deep neural networks. This paper studies the overparameterized linear models as a proxy for more complex models. We develop theoretical results describing the effect of various system parameters on the model's performance in an MTL setup. Specifically, we study the impact of model size, dataset size, and task similarity on the generalization error and knowledge transfer. Additionally, we present theoretical results to characterize the performance of replay-based CL models. Our results reveal the impact of buffer size and model capacity on the forgetting rate in a CL setup and help shed light on some of the state-of-the-art CL methods. Finally, through extensive empirical evaluations, we demonstrate that our theoretical findings are also applicable to deep neural networks, offering valuable guidance for designing MTL and CL models in practice.
Contrastive learning with the nearest neighbor has proved to be one of the most efficient self-supervised learning (SSL) techniques by utilizing the similarity of multiple instances within the same class. However, its efficacy is constrained as the nearest neighbor algorithm primarily identifies ``easy'' positive pairs, where the representations are already closely located in the embedding space. In this paper, we introduce a novel approach called Contrastive Learning with Synthetic Positives (CLSP) that utilizes synthetic images, generated by an unconditional diffusion model, as the additional positives to help the model learn from diverse positives. Through feature interpolation in the diffusion model sampling process, we generate images with distinct backgrounds yet similar semantic content to the anchor image. These images are considered ``hard'' positives for the anchor image, and when included as supplementary positives in the contrastive loss, they contribute to a performance improvement of over 2\% and 1\% in linear evaluation compared to the previous NNCLR and All4One methods across multiple benchmark datasets such as CIFAR10, achieving state-of-the-art methods. On transfer learning benchmarks, CLSP outperforms existing SSL frameworks on 6 out of 8 downstream datasets. We believe CLSP establishes a valuable baseline for future SSL studies incorporating synthetic data in the training process.
This study evaluates the effectiveness of deep learning models in classifying histopathological images for early and accurate detection of breast cancer. Eight advanced models, including ResNet-50, DenseNet-121, ResNeXt-50, Vision Transformer (ViT), GoogLeNet (Inception v3), EfficientNet, MobileNet, and SqueezeNet, were compared using a dataset of 277,524 image patches. The Vision Transformer (ViT) model, with its attention-based mechanisms, achieved the highest validation accuracy of 94%, outperforming conventional CNNs. The study demonstrates the potential of advanced machine learning methods to enhance precision and efficiency in breast cancer diagnosis in clinical settings.
We introduce LLaVA-MoD, a novel framework designed to enable the efficient training of small-scale Multimodal Language Models (s-MLLM) by distilling knowledge from large-scale MLLM (l-MLLM). Our approach tackles two fundamental challenges in MLLM distillation. First, we optimize the network structure of s-MLLM by integrating a sparse Mixture of Experts (MoE) architecture into the language model, striking a balance between computational efficiency and model expressiveness. Second, we propose a progressive knowledge transfer strategy to ensure comprehensive knowledge migration. This strategy begins with mimic distillation, where we minimize the Kullback-Leibler (KL) divergence between output distributions to enable the student model to emulate the teacher network's understanding. Following this, we introduce preference distillation via Direct Preference Optimization (DPO), where the key lies in treating l-MLLM as the reference model. During this phase, the s-MLLM's ability to discriminate between superior and inferior examples is significantly enhanced beyond l-MLLM, leading to a better student that surpasses its teacher, particularly in hallucination benchmarks. Extensive experiments demonstrate that LLaVA-MoD outperforms existing models across various multimodal benchmarks while maintaining a minimal number of activated parameters and low computational costs. Remarkably, LLaVA-MoD, with only 2B activated parameters, surpasses Qwen-VL-Chat-7B by an average of 8.8% across benchmarks, using merely 0.3% of the training data and 23% trainable parameters. These results underscore LLaVA-MoD's ability to effectively distill comprehensive knowledge from its teacher model, paving the way for the development of more efficient MLLMs. The code will be available on: https://github.com/shufangxun/LLaVA-MoD.
Achieving personalized intelligence at the edge with real-time learning capabilities holds enormous promise in enhancing our daily experiences and helping decision making, planning, and sensing. However, efficient and reliable edge learning remains difficult with current technology due to the lack of personalized data, insufficient hardware capabilities, and inherent challenges posed by online learning. Over time and across multiple developmental stages, the brain has evolved to efficiently incorporate new knowledge by gradually building on previous knowledge. In this work, we emulate the multiple stages of learning with digital neuromorphic technology that simulates the neural and synaptic processes of the brain using two stages of learning. First, a meta-training stage trains the hyperparameters of synaptic plasticity for one-shot learning using a differentiable simulation of the neuromorphic hardware. This meta-training process refines a hardware local three-factor synaptic plasticity rule and its associated hyperparameters to align with the trained task domain. In a subsequent deployment stage, these optimized hyperparameters enable fast, data-efficient, and accurate learning of new classes. We demonstrate our approach using event-driven vision sensor data and the Intel Loihi neuromorphic processor with its plasticity dynamics, achieving real-time one-shot learning of new classes that is vastly improved over transfer learning. Our methodology can be deployed with arbitrary plasticity models and can be applied to situations demanding quick learning and adaptation at the edge, such as navigating unfamiliar environments or learning unexpected categories of data through user engagement.
Language Language Models (LLMs) face safety concerns due to potential misuse by malicious users. Recent red-teaming efforts have identified adversarial suffixes capable of jailbreaking LLMs using the gradient-based search algorithm Greedy Coordinate Gradient (GCG). However, GCG struggles with computational inefficiency, limiting further investigations regarding suffix transferability and scalability across models and data. In this work, we bridge the connection between search efficiency and suffix transferability. We propose a two-stage transfer learning framework, DeGCG, which decouples the search process into behavior-agnostic pre-searching and behavior-relevant post-searching. Specifically, we employ direct first target token optimization in pre-searching to facilitate the search process. We apply our approach to cross-model, cross-data, and self-transfer scenarios. Furthermore, we introduce an interleaved variant of our approach, i-DeGCG, which iteratively leverages self-transferability to accelerate the search process. Experiments on HarmBench demonstrate the efficiency of our approach across various models and domains. Notably, our i-DeGCG outperforms the baseline on Llama2-chat-7b with ASRs of $43.9$ ($+22.2$) and $39.0$ ($+19.5$) on valid and test sets, respectively. Further analysis on cross-model transfer indicates the pivotal role of first target token optimization in leveraging suffix transferability for efficient searching.
Action recognition has become a hot topic in computer vision. However, the main applications of computer vision in video processing have focused on detection of relatively simple actions while complex events such as violence detection have been comparatively less investigated. This study focuses on the benchmarking of various deep learning techniques on a complex dataset. Next, a larger dataset is utilized to test the uplift from increasing volume of data. The dataset size increase from 500 to 1,600 videos resulted in a notable average accuracy improvement of 6% across four models.
The dynamics of burning plasmas in tokamaks are crucial for advancing controlled thermonuclear fusion. This study applies the NeuralPlasmaODE, a multi-region multi-timescale transport model, to simulate the complex energy transfer processes in ITER deuterium-tritium (D-T) plasmas. Our model captures the interactions between energetic alpha particles, electrons, and ions, which are vital for understanding phenomena such as thermal runaway instability. We employ neural ordinary differential equations (Neural ODEs) for the numerical derivation of diffusivity parameters, enabling precise modeling of energy interactions between different plasma regions. By leveraging transfer learning, we utilize model parameters derived from DIII-D experimental data, enhancing the efficiency and accuracy of our simulations without training from scratch. Applying this model to ITER's inductive and non-inductive operational scenarios, our results demonstrate that radiation and transport processes effectively remove excess heat from the core plasma, preventing thermal runaway instability. This study underscores the potential of machine learning in advancing our understanding and control of burning plasma dynamics in fusion reactors.
In supervised machine learning, privileged information (PI) is information that is unavailable at inference, but is accessible during training time. Research on learning using privileged information (LUPI) aims to transfer the knowledge captured in PI onto a model that can perform inference without PI. It seems that this extra bit of information ought to make the resulting model better. However, finding conclusive theoretical or empirical evidence that supports the ability to transfer knowledge using PI has been challenging. In this paper, we critically examine the assumptions underlying existing theoretical analyses and argue that there is little theoretical justification for when LUPI should work. We analyze LUPI methods and reveal that apparent improvements in empirical risk of existing research may not directly result from PI. Instead, these improvements often stem from dataset anomalies or modifications in model design misguidedly attributed to PI. Our experiments for a wide variety of application domains further demonstrate that state-of-the-art LUPI approaches fail to effectively transfer knowledge from PI. Thus, we advocate for practitioners to exercise caution when working with PI to avoid unintended inductive biases.
Deep learning techniques have revolutionized image classification by mimicking human cognition and automating complex decision-making processes. However, the deployment of AI systems in the wild, especially in high-security domains such as defence, is curbed by the lack of explainability of the model. To this end, eXplainable AI (XAI) is an emerging area of research that is intended to explore the unexplained hidden black box nature of deep neural networks. This paper explores the application of the eXplainable Artificial Intelligence (XAI) tool to interpret the underwater image classification results, one of the first works in the domain to the best of our knowledge. Our study delves into the realm of SONAR image classification using a custom dataset derived from diverse sources, including the Seabed Objects KLSG dataset, the camera SONAR dataset, the mine SONAR images dataset, and the SCTD dataset. An extensive analysis of transfer learning techniques for image classification using benchmark Convolutional Neural Network (CNN) architectures such as VGG16, ResNet50, InceptionV3, DenseNet121, etc. is carried out. On top of this classification model, a post-hoc XAI technique, viz. Local Interpretable Model-Agnostic Explanations (LIME) are incorporated to provide transparent justifications for the model's decisions by perturbing input data locally to see how predictions change. Furthermore, Submodular Picks LIME (SP-LIME) a version of LIME particular to images, that perturbs the image based on the submodular picks is also extensively studied. To this end, two submodular optimization algorithms i.e. Quickshift and Simple Linear Iterative Clustering (SLIC) are leveraged towards submodular picks. The extensive analysis of XAI techniques highlights interpretability of the results in a more human-compliant way, thus boosting our confidence and reliability.
In this report we investigate the use of the Tustin neural network architecture (Tustin-Net) for the identification of a physical rotary inverse pendulum. This physics-based architecture is of particular interest as it builds on the known relationship between velocities and positions. We here aim at discussing the advantages, limitations and performance of Tustin-Nets compared to first-principles grey-box models on a real physical apparatus, showing how, with a standard training procedure, the former can hardly achieve the same accuracy as the latter. To address this limitation, we present a training strategy based on transfer learning that yields Tustin-Nets that are competitive with the first-principles model, without requiring extensive knowledge of the setup as the latter.
The use of single-photon sources (SPSs) is central to numerous systems and devices proposed amidst a modern surge in quantum technology. However, manufacturing schemes remain imperfect, and single-photon emission purity must often be experimentally verified via interferometry. Such a process is typically slow and costly, which has motivated growing research into whether SPS quality can be more rapidly inferred from incomplete emission statistics. Hence, this study is a sequel to previous work that demonstrated significant uncertainty in the standard method of quality estimation, i.e. the least-squares fitting of a physically motivated function, and asks: can machine learning (ML) do better? The study leverages eight datasets obtained from measurements involving an exemplary quantum emitter, i.e. a single InGaAs/GaAs epitaxial quantum dot; these eight contexts predominantly vary in the intensity of the exciting laser. Specifically, via a form of `transfer learning', five ML models, three linear and two ensemble-based, are trained on data from seven of the contexts and tested on the eighth. Validation metrics quickly reveal that even a linear regressor can outperform standard fitting when it is tested on the same contexts it was trained on, but the success of transfer learning is less assured, even though statistical analysis, made possible by data augmentation, suggests its superiority as an early estimator. Accordingly, the study concludes by discussing future strategies for grappling with the problem of SPS context dissimilarity, e.g. feature engineering and model adaptation.
In this paper, we propose a novel loss function aimed at optimizing the binary flare prediction problem by embedding the intrinsic ordinal flare characteristics into the binary cross-entropy (BCE) loss function. This modification is intended to provide the model with better guidance based on the ordinal characteristics of the data and improve the overall performance of the models. For our experiments, we employ a ResNet34-based model with transfer learning to predict $\geq$M-class flares by utilizing the shape-based features of magnetograms of active region (AR) patches spanning from $-$90$^{\circ}$ to $+$90$^{\circ}$ of solar longitude as our input data. We use a composite skill score (CSS) as our evaluation metric, which is calculated as the geometric mean of the True Skill Score (TSS) and the Heidke Skill Score (HSS) to rank and compare our models' performance. The primary contributions of this work are as follows: (i) We introduce a novel approach to encode ordinality into a binary loss function showing an application to solar flare prediction, (ii) We enhance solar flare forecasting by enabling flare predictions for each AR across the entire solar disk, without any longitudinal restrictions, and evaluate and compare performance. (iii) Our candidate model, optimized with the proposed loss function, shows an improvement of $\sim$7%, $\sim$4%, and $\sim$3% for AR patches within $\pm$30$^\circ$, $\pm$60$^\circ$, and $\pm$90$^\circ$ of solar longitude, respectively in terms of CSS, when compared with standard BCE. Additionally, we demonstrate the ability to issue flare forecasts for ARs in near-limb regions (regions between $\pm$60$^{\circ}$ to $\pm$90$^{\circ}$) with a CSS=0.34 (TSS=0.50 and HSS=0.23), expanding the scope of AR-based models for solar flare prediction. This advances the reliability of solar flare forecasts, leading to more effective prediction capabilities.
Subtle semantic differences in retinal image and text data present great challenges for pre-training visual-language models. Moreover, false negative samples, i.e., image-text pairs having the same semantics but incorrectly regarded as negatives, disrupt the visual-language pre-training process and affect the model's learning ability. This work aims to develop a retinal foundation model, called ViLReF, by pre-training on a paired dataset comprising 451,956 retinal images and corresponding diagnostic text reports. In our vision-language pre-training strategy, we leverage expert knowledge to facilitate the extraction of labels and propose a novel constraint, the Weighted Similarity Coupling Loss, to adjust the speed of pushing sample pairs further apart dynamically within the feature space. Furthermore, we employ a batch expansion module with dynamic memory queues, maintained by momentum encoders, to supply extra samples and compensate for the vacancies caused by eliminating false negatives. Extensive experiments are conducted on multiple datasets for downstream classification and segmentation tasks. The experimental results demonstrate the powerful zero-shot and transfer learning capabilities of ViLReF, verifying the effectiveness of our pre-training strategy. Our ViLReF model is available at: https://github.com/T6Yang/ViLReF.