Moreover, we develop stochastic gradient MCMC inference that is scalable to very long multivariate count time series.

Much ofreinforcement learning theory isbuiltontopoforacles that arecomputationally hard to implement.

Collective action against algorithmic systems provides an opportunity for a small group of individuals to strategically manipulate their data to get specific outcomes, from classification to recommendation models. This effectiveness will invite more growth of this type of coordinated actions, both in the size and the number of distinct collectives. With a small group, however, coordination is key. Currently, there is no formal analysis of how coordination challenges within a collective can impact downstream outcomes, or how multiple collectives may affect each other's success. In this work, we aim to provide guarantees on the success of collective action in the presence of both coordination noise and multiple groups. Our insight is that data generated by either multiple collectives or by coordination noise can be viewed as originating from multiple data distributions. Using this framing, we derive bounds on the success of collective action. We conduct experiments to study the effects of noise on collective action. We find that sufficiently high levels of noise can reduce the success of collective action. In certain scenarios, large noise can sink a collective success rate from $100\%$ to just under $60\%$. We identify potential trade-offs between collective size and coordination noise; for example, a collective that is twice as big but with four times more noise experiencing worse outcomes than the smaller, more coordinated one. This work highlights the importance of understanding nuanced dynamics of strategic behavior in algorithmic systems.

Recent advances in text-to-image generative models have raised concerns about their potential to produce harmful content when provided with malicious input text prompts. To address this issue, two main approaches have emerged: (1) fine-tuning the model to unlearn harmful concepts and (2) training-free guidance methods that leverage negative prompts. However, we observe that combining these two orthogonal approaches often leads to marginal or even degraded defense performance. This observation indicates a critical incompatibility between two paradigms, which hinders their combined effectiveness. In this work, we address this issue by proposing a conceptually simple yet experimentally robust method: replacing the negative prompts used in training-free methods with implicit negative embeddings obtained through concept inversion. Our method requires no modification to either approach and can be easily integrated into existing pipelines. We experimentally validate its effectiveness on nudity and violence benchmarks, demonstrating consistent improvements in defense success rate while preserving the core semantics of input prompts.

Modern regression analyses are often undermined by covariate measurement error, misspecification of the regression model, and misspecification of the measurement error distribution. We present, to the best of our knowledge, the first Bayesian nonparametric framework targeting total robustness that tackles all three challenges in general nonlinear regression. The framework assigns a Dirichlet process prior to the latent co-variate-response distribution and updates it with posterior pseudo-samples of the latent covariates, thereby providing the Dirichlet process posterior with observation-informed latent inputs and yielding estimators that minimise the discrepancy between Dirichlet process realisations and the model-induced joint law. This design allows practitioners to (i) encode prior beliefs, (ii) choose between pseudo-sampling latent covariates or working directly with error-prone observations, and (iii) tune the influence of prior and data. We establish generalisation bounds that tighten whenever the prior or pseudo-sample generator aligns with the underlying data generating process, ensuring robustness without sacrificing consistency. A gradient-based algorithm enables efficient computations; simulations and two real-world studies show lower estimation error and reduced estimation sensitivity to misspecification compared to Bayesian and frequentist competitors. The framework, therefore, offers a practical and interpretable paradigm for trustworthy regression when data and models are jointly imperfect.

Group-fairness metrics (e.g., equalized odds) can vary sharply across resamples and are especially brittle under distribution shift, undermining reliable audits. We propose a Wasserstein distributionally robust framework that certifies worst-case group fairness over a ball of plausible test distributions centered at the empirical law. Our formulation unifies common group fairness notions via a generic conditional-probability functional and defines $\varepsilon$-Wasserstein Distributional Fairness ($\varepsilon$-WDF) as the audit target. Leveraging strong duality, we derive tractable reformulations and an efficient estimator (DRUNE) for $\varepsilon$-WDF. We prove feasibility and consistency and establish finite-sample certification guarantees for auditing fairness, along with quantitative bounds under smoothness and margin conditions. Across standard benchmarks and classifiers, $\varepsilon$-WDF delivers stable fairness assessments under distribution shift, providing a principled basis for auditing and certifying group fairness beyond observational data.

Curriculum learning, a training technique where data is presented to the model in order of example difficulty (e.g., from simpler to more complex documents), has shown limited success for pre-training language models. In this work, we investigate whether curriculum learning becomes competitive if we replace conventional human-centered difficulty metrics with one that more closely corresponds to example difficulty as observed during model training. Specifically, we experiment with sorting training examples by their \textit{training data influence}, a score which estimates the effect of individual training examples on the model's output. Models trained on our curricula are able to outperform ones trained in random order by over 10 percentage points in benchmarks, confirming that curriculum learning is beneficial for language model pre-training, as long as a more model-centric notion of difficulty is adopted.