The risk of language models reproducing copyrighted material from their training data has led to the development of various protective measures. Among these, inference-time strategies that impose constraints via post-processing have shown promise in addressing the complexities of copyright regulation. However, they often incur prohibitive computational costs or suffer from performance trade-offs. To overcome these limitations, we introduce Copyright-Protecting Model Fusion (CP-Fuse), a novel approach that combines models trained on disjoint sets of copyrighted material during inference. In particular, CP-Fuse adaptively aggregates the model outputs to minimize the reproduction of copyrighted content, adhering to a crucial balancing property to prevent the regurgitation of memorized data. Through extensive experiments, we show that CP-Fuse significantly reduces the reproduction of protected material without compromising the quality of text and code generation. Moreover, its post-hoc nature allows seamless integration with other protective measures, further enhancing copyright safeguards. Lastly, we show that CP-Fuse is robust against common techniques for extracting training data.

Practical and ethical constraints often require the use of observational data for causal inference, particularly in medicine and social sciences. Yet, observational datasets are prone to confounding, potentially compromising the validity of causal conclusions. While it is possible to correct for biases if the underlying causal graph is known, this is rarely a feasible ask in practical scenarios. A common strategy is to adjust for all available covariates, yet this approach can yield biased treatment effect estimates, especially when post-treatment or unobserved variables are present. We propose RAMEN, an algorithm that produces unbiased treatment effect estimates by leveraging the heterogeneity of multiple data sources without the need to know or learn the underlying causal graph. Notably, RAMEN achieves *doubly robust identification*: it can identify the treatment effect whenever the causal parents of the treatment or those of the outcome are observed, and the node whose parents are observed satisfies an invariance assumption. Empirical evaluations across synthetic, semi-synthetic, and real-world datasets show that our approach significantly outperforms existing methods.

Randomized experiments are the preferred approach for evaluating the effects of interventions, but they are costly and often yield estimates with substantial uncertainty. On the other hand, in silico experiments leveraging foundation models offer a cost-effective alternative that can potentially attain higher statistical precision. However, the benefits of in silico experiments come with a significant risk: statistical inferences are not valid if the models fail to accurately predict experimental responses to interventions. In this paper, we propose a novel approach that integrates the predictions from multiple foundation models with experimental data while preserving valid statistical inference. Our estimator is consistent and asymptotically normal, with asymptotic variance no larger than the standard estimator based on experimental data alone. Importantly, these statistical properties hold even when model predictions are arbitrarily biased. Empirical results across several randomized experiments show that our estimator offers substantial precision gains, equivalent to a reduction of up to 20\% in the sample size needed to match the same precision as the standard estimator based on experimental data alone.

Randomized trials are considered the gold standard for making informed decisions in medicine. However, they are often not representative of the patient population in clinical practice. Observational studies, on the other hand, cover a broader patient population but are prone to various biases. Thus, before using observational data for any downstream task, it is crucial to benchmark its treatment effect estimates against a randomized trial. We propose a novel strategy to benchmark observational studies on a subgroup level. First, we design a statistical test for the null hypothesis that the treatment effects – conditioned on a subset of relevant features – differ up to some tolerance value. Our test allows us to estimate an asymptotically valid lower bound on the maximum bias strength for any subgroup. We validate our lower bound in a real-world setting and show that it leads to conclusions that align with established medical knowledge.

We consider the estimation of the unknown hyperparameters for the problem of reconstructing a high-resolution image from multiple undersampled, shifted, degraded frames with subpixel displacement errors. We derive mathematical expressions for the iterative calculation of the maximum likelihood estimate (MLE) of the unknown hyperparameters given the low resolution observed images. Experimental results are presented for evaluating the accuracy of the proposed method.

In this paper, we consider the estimation of the unknown parameters for the problem of reconstructing a high-resolution image from multiple undersampled, shifted, degraded frames with subpixel displacement errors. We derive mathematical expressions for the iterative calculation of the maximum likelihood estimate of the unknown parameters given the low resolution observed images. These iterative procedures require the manipulation of block-semi circulant (BSC) matrices, that is, block matrices with circulant blocks. We show how these BSC matrices can be easily manipulated in order to calculate the unknown parameters. Finally the proposed method is tested on real and synthetic images.

The subband decomposition of a single channel image restoration problem is examined. The decomposition is carried out in the image model (prior model) in order to take into account the frequency activity of each band of the original image. The hyperparameters associated with each band together with the original image are rigorously estimated within the Bayesian framework. Finally, the proposed method is tested and compared with other methods on real images.

This paper deals with the simultaneous identification of the blur and the restoration of a noisy and blurred image. We propose the use of Dirichlet distributions to model our prior knowledge about the blurring function together with smoothness constraints on the restored image to solve the blind deconvolution problem. We show that the use of Dirichlet distributions offers a lot of flexibility in incorporating vague or very precise knowledge about the blurring process into the blind deconvolution process. The proposed MAP estimator offers additional flexibility in modeling the original image. Experimental results demonstrate the performance of the proposed algorithm.

We examine the use of compound Gauss Markov random fields (CGMRF) to restore severely blurred high range images. For this deblurring problem, the convergence of the simulated annealing (SA) and iterative conditional mode (ICM) algorithms has not been established. We propose two new iterative restoration algorithms which extend the classical SA and ICM approaches. Their convergence is established and they are tested on real and synthetic images.