Fine-tuned language models pose significant privacy risks, as they may memorize and expose sensitive information from their training data.Membership inference attacks (MIAs) provide a principled framework for auditing these risks, yet existing methods achieve limited detection rates, particularly at the low falsepositive thresholds required for practical privacy auditing.We present EZ-MIA, a membership inference attack that exploits a key observation: memorization manifests most strongly at error positions, specifically tokens where the model predicts incorrectly yet still shows elevated probability for training examples.We introduce the Error Zone (EZ) score, which measures the directional imbalance of probability shifts at error positions relative to a pretrained reference model.This principled statistic requires only two forward passes per query and no model training of any kind.On WikiText with GPT-2, EZ-MIA achieves 3.8× higher detection than the previous state-of-the-art under identical conditions (66.3% versus 17.5% true positive rate at 1% false positive rate), with near-perfect discrimination (AUC 0.98).At the stringent 0.1% FPR threshold critical for realworld auditing, we achieve 8× higher detection than prior work (14.0%versus 1.8%), requiring no reference model training.These gains extend to larger architectures: on AG News with Llama-2-7B, we achieve 3× higher detection (46.7% versus 15.8% TPR at 1% FPR).These results establish that privacy risks of finetuned language models are substantially greater than previously understood, with implications for both privacy auditing and deployment decisions.

2519

Dynamical systems governed by ordinary differential equations (ODEs) serve as models for a vast number of natural and social phenomena. In this work, we offer a fresh perspective on the classical problem of imputing missing time series data, whose underlying dynamics are assumed to be determined by ODEs. Specifically, we revisit ideas from amortized inference and neural operators, and propose a novel supervised learning framework for *zero-shot time series imputation*, through parametric functions satisfying some (hidden) ODEs. Our proposal consists of two components. First, a broad probability distribution over the space of ODE solutions, observation times and noise mechanisms, with which we generate a large, synthetic dataset of (hidden) ODE solutions, along with their noisy and sparse observations. Second, a neural recognition model that is trained *offline*, to map the generated time series onto the spaces of initial conditions and time derivatives of the (hidden) ODE solutions, which we then integrate to impute the missing data. We empirically demonstrate that *one and the same* (pretrained) recognition model can perform zero-shot imputation across 63 distinct time series with missing values, each sampled from widely different dynamical systems. Likewise, we demonstrate that it can perform zero-shot imputation of missing high-dimensional data in 10 vastly different settings, spanning human motion, air quality, traffic and electricity studies, as well as Navier-Stokes simulations — *without requiring any fine-tuning*. What is more, our proposal often outperforms state-of-the-art methods, which are trained on the target datasets. Our pretrained model, repository and tutorials are available online.

Stochastic differential equations (SDEs) describe dynamical systems where deterministic flows, governed by a drift function, are superimposed with random fluctuations, dictated by a diffusion function. The accurate estimation (*or discovery*) of these functions from data is a central problem in machine learning, with wide application across the natural and social sciences. Yet current solutions either rely heavily on prior knowledge of the dynamics or involve intricate training procedures. We introduce FIM-SDE (Foundation Inference Model for SDEs), a pretrained recognition model that delivers accurate *in-context* (or zero-shot) estimation of the drift and diffusion functions of *low-dimensional* SDEs, from noisy time series data, and allows rapid *finetuning* to target datasets. Leveraging concepts from amortized inference and neural operators, we (pre)train FIM-SDE in a supervised fashion to map a large set of noisy, discretely observed SDE paths onto the space of drift and diffusion functions. We demonstrate that FIM-SDE achieves robust *in-context* function estimation across a wide range of synthetic and real-world processes --- from canonical SDE systems (*e.g*., double-well dynamics or weakly perturbed Lorenz attractors) to stock price recordings and oil-price and wind-speed fluctuations --- while matching the performance of symbolic, Gaussian process and Neural SDE baselines trained on the target datasets. When *finetuned* to the target processes, we show that FIM-SDE consistently outperforms all these baselines.

Markov jump processes are continuous-time stochastic processes which describe dynamical systems evolving in discrete state spaces. These processes find wide application in the natural sciences and machine learning, but their inference is known to be far from trivial. In this work we introduce a methodology for *zero-shot inference* of Markov jump processes (MJPs), on bounded state spaces, from noisy and sparse observations, which consists of two components. First, a broad probability distribution over families of MJPs, as well as over possible observation times and noise mechanisms, with which we simulate a synthetic dataset of hidden MJPs and their noisy observations. Second, a neural recognition model that processes subsets of the simulated observations, and that is trained to output the initial condition and rate matrix of the target MJP in a supervised way. We empirically demonstrate that *one and the same* (pretrained) recognition model can infer, *in a zero-shot fashion*, hidden MJPs evolving in state spaces of different dimensionalities. Specifically, we infer MJPs which describe (i) discrete flashing ratchet systems, which are a type of Brownian motors, and the conformational dynamics in (ii) molecular simulations, (iii) experimental ion channel data and (iv) simple protein folding models. What is more, we show that our model performs on par with state-of-the-art models which are trained on the target datasets. Our pretrained model is available online.

Topic models and all their variants analyse text by learning meaningful representations through word co-occurrences. As pointed out by previous work, such models implicitly assume that the probability of a topic to be active and its proportion within each document are positively correlated. This correlation can be strongly detrimental in the case of documents created over time, simply because recent documents are likely better described by new and hence rare topics. In this work we leverage recent advances in neural variational inference and present an alternative neural approach to the dynamic Focused Topic Model. Indeed, we develop a neural model for topic evolution which exploits sequences of Bernoulli random variables in order to track the appearances of topics, thereby decoupling their activities from their proportions. We evaluate our model on three different datasets (the UN general debates, the collection of NeurIPS papers, and the ACL Anthology dataset) and show that it (i) outperforms state-of-the-art topic models in generalization tasks and (ii) performs comparably to them on prediction tasks, while employing roughly the same number of parameters, and converging about two times faster.

Ramses Sanchez, Lukas Conrads, Pascal Welke, Kostadin Cvejoski, Cesar Ojeda Marin. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Password generation techniques have recently been explored by leveraging deep-learning natural language processing (NLP) algorithms. Previous work has raised the state of the art for password guessing algorithms significantly, by approaching the problem using either variational autoencoders with CNN-based encoder and decoder architectures or transformer-based architectures (namely GPT2) for text generation. In this work we aim to combine both paradigms, introducing a novel architecture that leverages the expressive power of transformers with the natural sampling approach to text generation of variational autoencoders. We show how our architecture generates state-of-the-art results in password matching performance across multiple benchmark datasets.

Modern society is heavily dependent on large scale client-server systems with applications ranging from Internet and Communication Services to sophisticated logistics and deployment of goods. To maintain and improve such a system, a careful study of client and server dynamics is needed – e.g. response/service times, aver-age number of clients at given times, etc. To this end, one traditionally relies, within the queuing theory formalism,on parametric analysis and explicit distribution forms.However, parametric forms limit the model’s expressiveness and could struggle on extensively large datasets. We propose a novel data-driven approach towards queuing systems: the Deep Generative Service Times. Our methodology delivers a flexible and scalable model for service and response times. We leverage the representation capabilities of Recurrent Marked Point Processes for the temporal dynamics of clients, as well as Wasserstein Generative Adversarial Network techniques, to learn deep generative models which are able to represent complex conditional service time distributions. We provide extensive experimental analysis on both empirical and synthetic datasets, showing the effectiveness of the proposed models

Inspired by explanations of machine learning concepts in children’s books, we developed an approach to introduce supervised, unsupervised, and reinforcement learning using a block-based programming language in combination with the benefits of educational robotics. Instead of using blocks as high-end APIs to access AI cloud services or to reproduce the machine learning algorithms, we use them as a means to put the student “in the algorithm’s shoes.” We adapt the training of neural networks, Q-learning, and k-means algorithms to a design and format suitable for children and equip the students with hands-on tools for playful experimentation. The children learn about direct supervision by modifying the weights in the neural networks and immediately observing the effects on the simulated robot. Following the ideas of constructionism, they experience how the algorithms and underlying machine learning concepts work in practice. We conducted and evaluated this approach with students in primary, middle, and high school. All the age groups perceived the topics to be very easy to moderately hard to grasp. Younger students experienced direct supervision as challenging, whereas they found Q-learning and k-means algorithms much more accessible. Most high-school students could cope with all the topics without particular difficulties.

The problem of uncovering different dynamical regimes is of pivotal importance in time series analysis. Switching dynamical systems provide a solution for modeling physical phenomena whose time series data exhibit different dynamical modes. In this work we propose a novel variational RNN model for switching dynamics allowing for both non-Markovian and nonlinear dynamical behavior between and within dynamic modes. Attention mechanisms are provided to inform the switching distribution. We evaluate our model on synthetic and empirical datasets of diverse nature and successfully uncover different dynamical regimes and predict the switching dynamics.

In this paper we ask for the main factors that determine a classifier's decision making process and uncover such factors by studying latent codes produced by auto-encoding frameworks. To deliver an explanation of a classifier's behaviour, we propose a method that provides series of examples highlighting semantic differences between the classifier's decisions. These examples are generated through interpolations in latent space. We introduce and formalize the notion of a semantic stochastic path, as a suitable stochastic process defined in feature (data) space via latent code interpolations. We then introduce the concept of semantic Lagrangians as a way to incorporate the desired classifier's behaviour and find that the solution of the associated variational problem allows for highlighting differences in the classifier decision. Very importantly, within our framework the classifier is used as a black-box, and only its evaluation is required.

Deep neural network models represent the state-of-the-art methodologies for natural language processing. Here we build on top of these methodologies to incorporate temporal information and model how review data changes with time. Specifically, we use the dynamic representations of recurrent point process models, which encode the history of how business or service reviews are received in time, to generate instantaneous language models with improved prediction capabilities. Simultaneously, our methodologies enhance the predictive power of our point process models by incorporating summarized review content representations. We provide recurrent network and temporal convolution solutions for modeling the review content. We deploy our methodologies in the context of recommender systems, effectively characterizing the change in preference and taste of users as time evolves. Source code is available at [1].