--Point cloud place recognition (PCPR) determines the geo-location within a prebuilt map and plays a crucial role in geoscience and robotics applications such as autonomous driving, intelligent transportation, and augmented reality. In real-world large-scale deployments of a geographic positioning system, PCPR models must continuously acquire, update, and accumulate knowledge to adapt to diverse and dynamic environments, i.e., the ability known as continual learning (CL). However, existing PCPR models often suffer from catastrophic forgetting, leading to significant performance degradation in previously learned scenes when adapting to new environments or sensor types. This results in poor model scalability, increased maintenance costs, and system deployment difficulties, undermining the practicality of PCPR. T o address these issues, we propose LifelongPR, a novel continual learning framework for PCPR, which effectively extracts and fuses knowledge from sequential point cloud data. First, to alleviate the knowledge loss, we propose a replay sample selection method that dynamically allocates sample sizes according to each dataset's information quantity and selects spatially diverse samples for maximal representativeness. Second, to handle domain shifts, we design a prompt learning-based CL framework with a lightweight prompt module and a two-stage training strategy, enabling domain-specific feature adaptation while minimizing forgetting. Comprehensive experiments on large-scale public and self-collected datasets are conducted to validate the effectiveness of the proposed method. Compared with the state-of-the-art (SOT A) method, our method achieves 6.50% improvement in mIR @1, 7.96% improvement in mR @1, and an 8.95% reduction in F . LACE recognition is a foundational task in geoscience and robotics, enabling autonomous systems to determining their geo-locations within previously mapped environments by identifying revisited places [1, 2]. This study was supported by the National Natural Science Foundation Project (No. 42130105, No. 42201477, No. 42171431). Jianping Li is with the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore 639798.

Finetuning provides a scalable and cost-effective means of customizing language models for specific tasks or response styles, with greater reliability than prompting or in-context learning. In contrast, the conventional wisdom is that injecting knowledge via finetuning results in brittle performance and poor generalization. We argue that the dichotomy of "task customization" (e.g., instruction tuning) and "knowledge injection" (e.g., teaching new facts) is a distinction without a difference. We instead identify concrete factors that explain the heterogeneous effectiveness observed with finetuning. To this end, we conduct a large-scale experimental study of finetuning the frontier Gemini v1.5 model family on a spectrum of datasets that are artificially engineered to interpolate between the strengths and failure modes of finetuning. Our findings indicate that question-answer training data formats provide much stronger knowledge generalization than document/article-style training data, numerical information can be harder for finetuning to retain than categorical information, and models struggle to apply finetuned knowledge during multi-step reasoning even when trained on similar examples -- all factors that render "knowledge injection" to be especially difficult, even after controlling for considerations like data augmentation and information volume. On the other hand, our findings also indicate that it is not fundamentally more difficult to finetune information about a real-world event than information about what a model's writing style should be.

This paper develops new information quantities. These quantities are derived from the Hyvarinen loss and are shown to be related to Fisher divergence. The behavior of some of these quantities is similar in part to classical information quantities such as entropy, mutual information, KL divergence etc. The former quantities are said to be faster to compute and more robust than the latter. In this paper, these quantities are derived, studied and used to derive several learning algorithms, most notably a version of the Chow-Liu algorithm based on these quantities.

The concepts of conditional mutual information (CMI) and normalized conditional mutual information (NCMI) are introduced to measure the concentration and separation performance of a classification deep neural network (DNN) in the output probability distribution space of the DNN, where CMI and the ratio between CMI and NCMI represent the intra-class concentration and inter-class separation of the DNN, respectively. By using NCMI to evaluate popular DNNs pretrained over ImageNet in the literature, it is shown that their validation accuracies over ImageNet validation data set are more or less inversely proportional to their NCMI values. Based on this observation, the standard deep learning (DL) framework is further modified to minimize the standard cross entropy function subject to an NCMI constraint, yielding CMI constrained deep learning (CMIC-DL). A novel alternating learning algorithm is proposed to solve such a constrained optimization problem. Extensive experiment results show that DNNs trained within CMIC-DL outperform the state-of-the-art models trained within the standard DL and other loss functions in the literature in terms of both accuracy and robustness against adversarial attacks. In addition, visualizing the evolution of learning process through the lens of CMI and NCMI is also advocated.

Recently, information theoretic analysis has become a popular framework for understanding the generalization behavior of deep neural networks. It allows a direct analysis for stochastic gradient/Langevin descent (SGD/SGLD) learning algorithms without strong assumptions such as Lipschitz or convexity conditions. However, the current generalization error bounds within this framework are still far from optimal, while substantial improvements on these bounds are quite challenging due to the intractability of high-dimensional information quantities. To address this issue, we first propose a novel information theoretical measure: kernelized Renyi's entropy, by utilizing operator representation in Hilbert space. It inherits the properties of Shannon's entropy and can be effectively calculated via simple random sampling, while remaining independent of the input dimension. We then establish the generalization error bounds for SGD/SGLD under kernelized Renyi's entropy, where the mutual information quantities can be directly calculated, enabling evaluation of the tightness of each intermediate step. We show that our information-theoretical bounds depend on the statistics of the stochastic gradients evaluated along with the iterates, and are rigorously tighter than the current state-of-the-art (SOTA) results. The theoretical findings are also supported by large-scale empirical studies1.

Recently proposed methods in data subset selection, that is active learning and active sampling, use Fisher information, Hessians, similarity matrices based on gradients, and gradient lengths to estimate how informative data is for a model's training. Are these different approaches connected, and if so, how? We revisit the fundamentals of Bayesian optimal experiment design and show that these recently proposed methods can be understood as approximations to information-theoretic quantities: among them, the mutual information between predictions and model parameters, known as expected information gain or BALD in machine learning, and the mutual information between predictions of acquisition candidates and test samples, known as expected predictive information gain. We develop a comprehensive set of approximations using Fisher information and observed information and derive a unified framework that connects seemingly disparate literature. Although Bayesian methods are often seen as separate from non-Bayesian ones, the sometimes fuzzy notion of "informativeness" expressed in various non-Bayesian objectives leads to the same couple of information quantities, which were, in principle, already known by Lindley (1956) and MacKay (1992).

Bottleneck problems are an important class of optimization problems that have recently gained increasing attention in the domain of machine learning and information theory. They are widely used in generative models, fair machine learning algorithms, design of privacy-assuring mechanisms, and appear as information-theoretic performance bounds in various multi-user communication problems. In this work, we propose a general family of optimization problems, termed as complexity-leakage-utility bottleneck (CLUB) model, which (i) provides a unified theoretical framework that generalizes most of the state-of-the-art literature for the information-theoretic privacy models, (ii) establishes a new interpretation of the popular generative and discriminative models, (iii) constructs new insights to the generative compression models, and (iv) can be used in the fair generative models. We first formulate the CLUB model as a complexity-constrained privacy-utility optimization problem. We then connect it with the closely related bottleneck problems, namely information bottleneck (IB), privacy funnel (PF), deterministic IB (DIB), conditional entropy bottleneck (CEB), and conditional PF (CPF).

The information bottleneck (IB) principle offers both a mechanism to explain how deep neural networks train and generalize, as well as a regularized objective with which to train models. However, multiple competing objectives have been proposed based on this principle, and the information-theoretic quantities in these objectives are difficult to compute for large deep neural networks. This, in turn, limits their use as a training objective. In this work, we review these quantities, compare and unify previously proposed objectives and relate them to surrogate objectives more friendly to optimization. We find that these surrogate objectives allow us to apply the information bottleneck to modern neural network architectures. We demonstrate our insights on Permutation-MNIST, MNIST and CIFAR10.

Despite their great success in practical applications, there is still a lack of theoretical and systematic methods to analyze deep neural networks. In this paper, we illustrate an advanced information theoretic methodology to understand the dynamics of learning and the design of autoencoders, a special type of deep learning architectures that resembles a communication channel. By generalizing the information plane to any cost function, and inspecting the roles and dynamics of different layers using layer-wise information quantities, we emphasize the role that mutual information plays in quantifying learning from data. We further propose and also experimentally validate, for mean square error training, two hypotheses regarding the layer-wise flow of information and intrinsic dimensionality of the bottleneck layer, using respectively the data processing inequality and the identification of a bifurcation point in the information plane that is controlled by the given data. Our observations have direct impact on the optimal design of autoencoders, the design of alternative feedforward training methods, and even in the problem of generalization.

The types of web data vary in terms of information quantity and quality. For example, some pages contain numerous texts, whereas some others contain few texts; some web videos are in high resolution, whereas some other web videos are in low resolution. As a consequence, the quality of extracted features from different web data may also vary greatly. Existing learning algorithms on web data classification usually ignore the variations of information quality or quantity. In this paper, the information quantity and quality of web data are described by quality-related factors such as text length and image quantity, and a new learning method is proposed to train classifiers based on quality-related factors. The method divides training data into subsets according to the clustering results of quality-related factors and then trains classifiers by using a multi-task learning strategy for each subset. Experimental results indicate that the quality-related factors are useful in web data classification, and the proposed method outperforms conventional algorithms that do not consider information quantity and quality.