ScenarioBench is a policy-grounded, trace-aware benchmark for evaluating Text-to-SQL and retrieval-augmented generation in compliance contexts. Each YAML scenario includes a no-peek gold-standard package with the expected decision, a minimal witness trace, the governing clause set, and the canonical SQL, enabling end-to-end scoring of both what a system decides and why. Systems must justify outputs using clause IDs from the same policy canon, making explanations falsifiable and audit-ready. The evaluator reports decision accuracy, trace quality (completeness, correctness, order), retrieval effectiveness, SQL correctness via result-set equivalence, policy coverage, latency, and an explanation-hallucination rate. A normalized Scenario Difficulty Index (SDI) and a budgeted variant (SDI-R) aggregate results while accounting for retrieval difficulty and time. Compared with prior Text-to-SQL or KILT/RAG benchmarks, ScenarioBench ties each decision to clause-level evidence under strict grounding and no-peek rules, shifting gains toward justification quality under explicit time budgets.

The experimental results appear to very good compared to other CS-MRI approaches, since the proposed ADMM-Net offers both high quality and fast runtime, whereas the competing methods either have only fast runtime or good reconstruction quality. I think generalization of the learned network is an important point that deserves more emphasis. It is only briefly mentioned (l. Would this kind of generalization ability be useful in practice if this were to be used with an actual MRI machine? Overall, the proposed unrolled inference approach is similar to [19], which uses a similar data prior with similar inference (but without Lagrange multipliers) for image denoising and deconvolution.

Push-Relabel is one of the most celebrated network flow algorithms. Maintaining a pre-flow that saturates a cut, it enjoys better theoretical and empirical running time than other flow algorithms, such as Ford-Fulkerson. In practice, Push-Relabel is even faster than what theoretical guarantees can promise, in part because of the use of good heuristics for seeding and updating the iterative algorithm. However, it remains unclear how to run Push-Relabel on an arbitrary initialization that is not necessarily a pre-flow or cut-saturating. We provide the first theoretical guarantees for warm-starting Push-Relabel with a predicted flow, where our learning-augmented version benefits from fast running time when the predicted flow is close to an optimal flow, while maintaining robust worst-case guarantees. Interestingly, our algorithm uses the gap relabeling heuristic, which has long been employed in practice, even though prior to our work there was no rigorous theoretical justification for why it can lead to run-time improvements. We then provide experiments that show our warm-started Push-Relabel also works well in practice.

In this work, we approach the problem of multi-task submodular optimization with the perspective of local distributional robustness, within the neighborhood of a reference distribution which assigns an importance score to each task. We initially propose to introduce a regularization term which makes use of the relative entropy to the standard multi-task objective. We then demonstrate through duality that this novel formulation itself is equivalent to the maximization of a submodular function, which may be efficiently carried out through standard greedy selection methods. This approach bridges the existing gap in the optimization of performance-robustness trade-offs in multi-task subset selection. To numerically validate our theoretical results, we test the proposed method in two different setting, one involving the selection of satellites in low Earth orbit constellations in the context of a sensor selection problem, and the other involving an image summarization task using neural networks. Our method is compared with two other algorithms focused on optimizing the performance of the worst-case task, and on directly optimizing the performance on the reference distribution itself. We conclude that our novel formulation produces a solution that is locally distributional robust, and computationally inexpensive.

Sigmoid function: The sigmoid function is a widely used activation function that maps the input to a value between 0 and 1. It is defined as follows: sigmoid(x) 1 / (1 exp(-x)). It is useful for binary classification tasks where the output is a probability. However, it has a tendency to saturate, meaning that for large values of x, the derivative becomes very small, which can slow down training. To use sigmoid function in a neural network, you can use the sigmoid activation function provided by most deep learning frameworks. ReLU function: The rectified linear unit (ReLU) function is another popular activation function that is widely used in deep neural networks.

Training deep neural networks can be a challenging task, especially for very deep models. A major part of this difficulty is due to the instability of the gradients computed via backpropagation. In this post, we will learn how to create a self-normalizing deep feed-forward neural network using Keras. This will solve the gradient instability issue, speeding up training convergence, and improving model performance. Disclaimer: This article is a brief summary with focus on implementation.

In artificial neural networks, activation function plays an important role in determining the output of the neuron. To make it sound more realistic, we can simply compare the activation function to the biological neurons which fire the signal to other connected neurons. Basically, the activation function maps an output value between the value 0 and 1. Guess Why?, Because the neuron has a value which is not confined to a scale, making it unclear when to fire. Thus this value from neurons is mapped to a scale to understand as in when to fire (as in case of biological neurons). Where w is the weight of a neuron, x is an input and b is bias value.

Deep Learning (DL) models are revolutionizing the business and technology world with jaw-dropping performances in one application area after another -- image classification, object detection, object tracking, pose recognition, video analytics, synthetic picture generation -- just to name a few. However, they are like anything but classical Machine Learning (ML) algorithms/techniques. DL models use millions of parameters and create extremely complex and highly nonlinear internal representations of the images or datasets that are fed to these models. Whereas for the classical ML, domain experts and data scientists often have to write hand-crafted algorithms to extract and represent high-dimensional features from the raw data, deep learning models, on the other hand, automatically extracts and work on these complex features. A lot of theory and mathematical machines behind the classical ML (regression, support vector machines, etc.) were developed with linear models in mind. However, practical real-life problems are often nonlinear in nature and therefore cannot be effectively solved using those ML methods.