The branch-and-cut algorithm is the method of choice to solve large scale integer programming problems in practice. A key ingredient of branch-and-cut is the use of cutting planes which are derived constraints that reduce the search space for an optimal solution. Selecting effective cutting planes to produce small branch-and-cut trees is a critical challenge in the branch-and-cut algorithm. Recent advances have employed a data-driven approach to select optimal cutting planes from a parameterized family, aimed at reducing the branch-and-bound tree size (in expectation) for a given distribution of integer programming instances. We extend this idea to the selection of the best cut generating function (CGF), which is a tool in the integer programming literature for generating a wide variety of cutting planes that generalize the well-known Gomory Mixed-Integer (GMI) cutting planes. We provide rigorous sample complexity bounds for the selection of an effective CGF from certain parameterized families that provably performs well for any specified distribution on the problem instances. Our empirical results show that the selected CGF can outperform the GMI cuts for certain distributions. Additionally, we explore the sample complexity of using neural networks for instance-dependent CGF selection.

Training neural networks which are robust to adversarial attacks remains an important problem in deep learning, especially as heavily overparameterized models are adopted in safety-critical settings. Drawing from recent work which reformulates the training problems for two-layer ReLU and polynomial activation networks as convex programs, we devise a convex semidefinite program (SDP) for adversarial training of polynomial activation networks via the S-procedure. We also derive a convex SDP to compute the minimum distance from a correctly classified example to the decision boundary of a polynomial activation network. Adversarial training for two-layer ReLU activation networks has been explored in the literature, but, in contrast to prior work, we present a scalable approach which is compatible with standard machine libraries and GPU acceleration. The adversarial training SDP for polynomial activation networks leads to large increases in robust test accuracy against $\ell^\infty$ attacks on the Breast Cancer Wisconsin dataset from the UCI Machine Learning Repository. For two-layer ReLU networks, we leverage our scalable implementation to retrain the final two fully connected layers of a Pre-Activation ResNet-18 model on the CIFAR-10 dataset. Our 'robustified' model achieves higher clean and robust test accuracies than the same architecture trained with sharpness-aware minimization.

We introduce the Riskman ontology & shapes for representing and analysing information about risk management for medical devices. Risk management is concerned with taking necessary precautions so a medical device does not cause harms for users or the environment. To date, risk management documentation is submitted to notified bodies (for certification) in the form of semi-structured natural language text. We propose to use classes from the Riskman ontology to logically model risk management documentation, and to use the included SHACL constraints to check for syntactic completeness and conformity to relevant standards. In particular, the ontology is modelled after ISO 14971 and the recently published VDE Spec 90025. Our proposed methodology has the potential to save many person-hours for both manufacturers (when creating risk management documentation) as well as notified bodies (when assessing submitted applications for certification), and thus offers considerable benefits for healthcare and, by extension, society as a whole.

Traditional computer-aided synthesis planning (CASP) methods rely on iterative single-step predictions, leading to exponential search space growth that limits efficiency and scalability. We introduce a transformer-based model that directly generates multi-step synthetic routes as a single string by conditionally predicting each molecule based on all preceding ones. The model accommodates specific conditions such as the desired number of steps and starting materials, outperforming state-of-the-art methods on the PaRoutes dataset with a 2.2x improvement in Top-1 accuracy on the n$_1$ test set and a 3.3x improvement on the n$_5$ test set. It also successfully predicts routes for FDA-approved drugs not included in the training data, showcasing its generalization capabilities. While the current suboptimal diversity of the training set may impact performance on less common reaction types, our approach presents a promising direction towards fully automated retrosynthetic planning.

We present a new challenge to examine whether large language models understand social norms. In contrast to existing datasets, our dataset requires a fundamental understanding of social norms to solve. Our dataset features the largest set of social norm skills, consisting of 402 skills and 12,383 questions covering a wide set of social norms ranging from opinions and arguments to culture and laws. We design our dataset according to the K-12 curriculum. This enables the direct comparison of the social understanding of large language models to humans, more specifically, elementary students. While prior work generates nearly random accuracy on our benchmark, recent large language models such as GPT3.5-Turbo and LLaMA2-Chat are able to improve the performance significantly, only slightly below human performance. We then propose a multi-agent framework based on large language models to improve the models' ability to understand social norms. This method further improves large language models to be on par with humans. Given the increasing adoption of large language models in real-world applications, our finding is particularly important and presents a unique direction for future improvements.

This paper introduces xRAG, an innovative context compression method tailored for retrieval-augmented generation. xRAG reinterprets document embeddings in dense retrieval--traditionally used solely for retrieval--as features from the retrieval modality. By employing a modality fusion methodology, xRAG seamlessly integrates these embeddings into the language model representation space, effectively eliminating the need for their textual counterparts and achieving an extreme compression rate. In xRAG, the only trainable component is the modality bridge, while both the retriever and the language model remain frozen. This design choice allows for the reuse of offline-constructed document embeddings and preserves the plug-and-play nature of retrieval augmentation. Experimental results demonstrate that xRAG achieves an average improvement of over 10% across six knowledge-intensive tasks, adaptable to various language model backbones, ranging from a dense 7B model to an 8x7B Mixture of Experts configuration. xRAG not only significantly outperforms previous context compression methods but also matches the performance of uncompressed models on several datasets, while reducing overall FLOPs by a factor of 3.53. Our work pioneers new directions in retrieval-augmented generation from the perspective of multimodality fusion, and we hope it lays the foundation for future efficient and scalable retrieval-augmented systems

Large language models (LLMs) store extensive factual knowledge, but the mechanisms behind how they store and express this knowledge remain unclear. The Knowledge Neuron (KN) thesis is a prominent theory for explaining these mechanisms. This theory is based on the knowledge localization (KL) assumption, which suggests that a fact can be localized to a few knowledge storage units, namely knowledge neurons. However, this assumption may be overly strong regarding knowledge storage and neglects knowledge expression mechanisms. Thus, we re-examine the KL assumption and confirm the existence of facts that do not adhere to it from both statistical and knowledge modification perspectives. Furthermore, we propose the Query Localization (QL) assumption. (1) Query-KN Mapping: The localization results are associated with the query rather than the fact. (2) Dynamic KN Selection: The attention module contributes to the selection of KNs for answering a query. Based on this, we further propose the Consistency-Aware KN modification method, which improves the performance of knowledge modification. We conduct 39 sets of experiments, along with additional visualization experiments, to rigorously validate our conclusions.

Leader election is one of the fundamental and well-studied problems in distributed computing. In this paper, we initiate the study of leader election using mobile agents. Suppose $n$ agents are positioned initially arbitrarily on the nodes of an arbitrary, anonymous, $n$-node, $m$-edge graph $G$. The agents relocate themselves autonomously on the nodes of $G$ and elect an agent as a leader such that the leader agent knows it is a leader and the other agents know they are not leaders. The objective is to minimize time and memory requirements. Following the literature, we consider the synchronous setting in which each agent performs its operations synchronously with others and hence the time complexity can be measured in rounds. The quest in this paper is to provide solutions without agents knowing any graph parameter, such as $n$, a priori. We first establish that, without agents knowing any graph parameter a priori, there exists a deterministic algorithm to elect an agent as a leader in $O(m)$ rounds with $O(n\log n)$ bits at each agent. Using this leader election result, we develop a deterministic algorithm for agents to construct a minimum spanning tree of $G$ in $O(m+n\log n)$ rounds using $O(n \log n)$ bits memory at each agent, without agents knowing any graph parameter a priori. Finally, using the same leader election result, we provide improved time/memory results for other fundamental distributed graph problems, namely, gathering, maximal independent set, and minimal dominating sets, removing the assumptions on agents knowing graph parameters a priori.

We present ChatScene, a Large Language Model (LLM)-based agent that leverages the capabilities of LLMs to generate safety-critical scenarios for autonomous vehicles. Given unstructured language instructions, the agent first generates textually described traffic scenarios using LLMs. These scenario descriptions are subsequently broken down into several sub-descriptions for specified details such as behaviors and locations of vehicles. The agent then distinctively transforms the textually described sub-scenarios into domain-specific languages, which then generate actual code for prediction and control in simulators, facilitating the creation of diverse and complex scenarios within the CARLA simulation environment. A key part of our agent is a comprehensive knowledge retrieval component, which efficiently translates specific textual descriptions into corresponding domain-specific code snippets by training a knowledge database containing the scenario description and code pairs. Extensive experimental results underscore the efficacy of ChatScene in improving the safety of autonomous vehicles. For instance, the scenarios generated by ChatScene show a 15% increase in collision rates compared to state-of-the-art baselines when tested against different reinforcement learning-based ego vehicles. Furthermore, we show that by using our generated safety-critical scenarios to fine-tune different RL-based autonomous driving models, they can achieve a 9% reduction in collision rates, surpassing current SOTA methods. ChatScene effectively bridges the gap between textual descriptions of traffic scenarios and practical CARLA simulations, providing a unified way to conveniently generate safety-critical scenarios for safety testing and improvement for AVs.

Scaling laws for large language models (LLMs) have provided useful guidance on how to train ever larger models for predictable performance gains. Time series forecasting shares a similar sequential structure to language, and is amenable to large-scale transformer architectures. Here we show that foundational decoder-only time series transformer models exhibit analogous scaling-behavior to LLMs, while architectural details (aspect ratio and number of heads) have a minimal effect over broad ranges. We assemble a large corpus of heterogenous time series data on which to train, and establish, for the first time, power-law scaling relations with respect to parameter count, dataset size, and training compute, spanning five orders of magnitude.