Privacy concerns have led to a surge in the creation of synthetic datasets, with diffusion models emerging as a promising avenue. Although prior studies have performed empirical evaluations on these models, there has been a gap in providing a mathematical characterization of their privacy-preserving capabilities. To address this, we present the pioneering theoretical exploration of the privacy preservation inherent in discrete diffusion models (DDMs) for discrete dataset generation. Focusing on per-instance differential privacy (pDP), our framework elucidates the potential privacy leakage for each data point in a given training dataset, offering insights into how the privacy loss of each point correlates with the dataset's distribution. Our bounds also show that training with $s$-sized data points leads to a surge in privacy leakage from $(\epsilon, O(\frac{1}{s^2\epsilon}))$-pDP to $(\epsilon, O(\frac{1}{s\epsilon}))$-pDP of the DDM during the transition from the pure noise to the synthetic clean data phase, and a faster decay in diffusion coefficients amplifies the privacy guarantee. Finally, we empirically verify our theoretical findings on both synthetic and real-world datasets.

This paper introduces Bayesian Flow Networks (BFNs), a new class of generative model in which the parameters of a set of independent distributions are modified with Bayesian inference in the light of noisy data samples, then passed as input to a neural network that outputs a second, interdependent distribution. Starting from a simple prior and iteratively updating the two distributions yields a generative procedure similar to the reverse process of diffusion models; however it is conceptually simpler in that no forward process is required. Discrete and continuous-time loss functions are derived for continuous, discretised and discrete data, along with sample generation procedures. Notably, the network inputs for discrete data lie on the probability simplex, and are therefore natively differentiable, paving the way for gradient-based sample guidance and few-step generation in discrete domains such as language modelling. The loss function directly optimises data compression and places no restrictions on the network architecture. In our experiments BFNs achieve competitive log-likelihoods for image modelling on dynamically binarized MNIST and CIFAR-10, and outperform all known discrete diffusion models on the text8 character-level language modelling task.

Pervasive across diverse domains, stochastic systems exhibit fluctuations in processes ranging from molecular dynamics to climate phenomena. The Langevin equation has served as a common mathematical model for studying such systems, enabling predictions of their temporal evolution and analyses of thermodynamic quantities, including absorbed heat, work done on the system, and entropy production. However, inferring the Langevin equation from observed trajectories remains challenging, particularly for nonlinear and high-dimensional systems. In this study, we present a comprehensive framework that employs Bayesian neural networks for inferring Langevin equations in both overdamped and underdamped regimes. Our framework first provides the drift force and diffusion matrix separately and then combines them to construct the Langevin equation. By providing a distribution of predictions instead of a single value, our approach allows us to assess prediction uncertainties, which can prevent potential misunderstandings and erroneous decisions about the system. We demonstrate the effectiveness of our framework in inferring Langevin equations for various scenarios including a neuron model and microscopic engine, highlighting its versatility and potential impact.

Kernel methods are widely used in machine learning, especially for classification problems. However, the theoretical analysis of kernel classification is still limited. This paper investigates the statistical performances of kernel classifiers. With some mild assumptions on the conditional probability $\eta(x)=\mathbb{P}(Y=1\mid X=x)$, we derive an upper bound on the classification excess risk of a kernel classifier using recent advances in the theory of kernel regression. We also obtain a minimax lower bound for Sobolev spaces, which shows the optimality of the proposed classifier. Our theoretical results can be extended to the generalization error of overparameterized neural network classifiers. To make our theoretical results more applicable in realistic settings, we also propose a simple method to estimate the interpolation smoothness of $2\eta(x)-1$ and apply the method to real datasets.

In such cases, transferring knowledge from the source domain becomes crucial, particularly because the Machine learning has made substantial advancements in limited initial data in the target domain may be insufficient recent decades, with its applications spanning a wide range of for effective learning. The extensive and diverse information fields such as image and speech recognition, natural language available from the source domains can significantly compensate processing, and autonomous driving. Despite these achievements, for this shortfall, providing a foundational knowledge base machine learning in biomedicine faces significant challenges, that the model can build upon as more target domain data particularly in data collection. The acquisition of labeled becomes available. Therefore, the efficiency and effectiveness data can be very costly or even unfeasible due to factors of learning in the target domain are greatly enhanced by the like ethical considerations, patient privacy, and the scarcity transferred knowledge from the source domains. of certain diseases. These challenges have led researchers to Online transfer learning entails leveraging knowledge from increasingly rely on utilizing data from related domains that a static source domain and applying it to an ongoing, evolving have a more abundant supply of data.

Causal discovery, the task of inferring causal structure from data, promises to accelerate scientific research, inform policy making, and more. However, the per-dataset nature of existing causal discovery algorithms renders them slow, data hungry, and brittle. Inspired by foundation models, we propose a causal discovery framework where a deep learning model is pretrained to resolve predictions from classical discovery algorithms run over smaller subsets of variables. This method is enabled by the observations that the outputs from classical algorithms are fast to compute for small problems, informative of (marginal) data structure, and their structure outputs as objects remain comparable across datasets. Our method achieves state-of-the-art performance on synthetic and realistic datasets, generalizes to data generating mechanisms not seen during training, and offers inference speeds that are orders of magnitude faster than existing models.

In this work, we review research studies that combine Reinforcement Learning (RL) and Large Language Models (LLMs), two areas that owe their momentum to the development of deep neural networks. We propose a novel taxonomy of three main classes based on the way that the two model types interact with each other. The first class, RL4LLM, includes studies where RL is leveraged to improve the performance of LLMs on tasks related to Natural Language Processing. L4LLM is divided into two sub-categories depending on whether RL is used to directly fine-tune an existing LLM or to improve the prompt of the LLM. In the second class, LLM4RL, an LLM assists the training of an RL model that performs a task that is not inherently related to natural language. We further break down LLM4RL based on the component of the RL training framework that the LLM assists or replaces, namely reward shaping, goal generation, and policy function. Finally, in the third class, RL+LLM, an LLM and an RL agent are embedded in a common planning framework without either of them contributing to training or fine-tuning of the other. We further branch this class to distinguish between studies with and without natural language feedback. We use this taxonomy to explore the motivations behind the synergy of LLMs and RL and explain the reasons for its success, while pinpointing potential shortcomings and areas where further research is needed, as well as alternative methodologies that serve the same goal.

Ecological rationality refers to the notion that humans are rational agents adapted to their environment. However, testing this theory remains challenging due to two reasons: the difficulty in defining what tasks are ecologically valid and building rational models for these tasks. In this work, we demonstrate that large language models can generate cognitive tasks, specifically category learning tasks, that match the statistics of real-world tasks, thereby addressing the first challenge. We tackle the second challenge by deriving rational agents adapted to these tasks using the framework of meta-learning, leading to a class of models called ecologically rational meta-learned inference (ERMI). ERMI quantitatively explains human data better than seven other cognitive models in two different experiments. It additionally matches human behavior on a qualitative level: (1) it finds the same tasks difficult that humans find difficult, (2) it becomes more reliant on an exemplar-based strategy for assigning categories with learning, and (3) it generalizes to unseen stimuli in a human-like way. Furthermore, we show that ERMI's ecologically valid priors allow it to achieve state-of-the-art performance on the OpenML-CC18 classification benchmark.

The expected loss is an upper bound to the model generalization error which admits robust PAC-Bayes bounds for learning. However, loss minimization is known to ignore misspecification, where models cannot exactly reproduce observations. This leads to significant underestimates of parameter uncertainties in the large data, or underparameterized, limit. We analyze the generalization error of near-deterministic, misspecified and underparametrized surrogate models, a regime of broad relevance in science and engineering. We show posterior distributions must cover every training point to avoid a divergent generalization error and derive an ensemble {ansatz} that respects this constraint, which for linear models incurs minimal overhead. The efficient approach is demonstrated on model problems before application to high dimensional datasets in atomistic machine learning. Parameter uncertainties from misspecification survive in the underparametrized limit, giving accurate prediction and bounding of test errors.

Bayesian optimisation requires fitting a Gaussian process model, which in turn requires specifying hyperparameters - most of the theoretical literature assumes those hyperparameters are known. The commonly used maximum likelihood estimator for hyperparameters of the Gaussian process is consistent only if the data fills the space uniformly, which does not have to be the case in Bayesian optimisation. Since no guarantees exist regarding the correctness of hyperparameter estimation, and those hyperparameters can significantly affect the Gaussian process fit, theoretical analysis of Bayesian optimisation with unknown hyperparameters is very challenging. Previously proposed algorithms with the no-regret property were only able to handle the special case of unknown lengthscales, reproducing kernel Hilbert space norm and applied only to the frequentist case. We propose a novel algorithm, HE-GP-UCB, which is the first algorithm enjoying the no-regret property in the case of unknown hyperparameters of arbitrary form, and which supports both Bayesian and frequentist settings. Our proof idea is novel and can easily be extended to other variants of Bayesian optimisation. We show this by extending our algorithm to the adversarially robust optimisation setting under unknown hyperparameters. Finally, we empirically evaluate our algorithm on a set of toy problems and show that it can outperform the maximum likelihood estimator.