"I lost count after 100," the 44-year-old, camouflage-clad instructor told Al Jazeera while observing three cadets of his drone flight school pilot their buzzing aircraft over a withering meadow just outside Kyiv. Sitting at a plastic table littered with tools and batteries, the cadets with their joysticks and goggle cameras looked geeky and harmless. During their Saturday morning drill, each of them took turns flying a drone whose camera allows first-person views of the flight. Time after time after time, the cadets learned how to manoeuvre their drones by flying them through two loops stuck into the wet ground. The drones often fell with a whiz after touching a loop or a bush, losing a red plastic propeller or a leg that had to be found in the wet grass and reattached.

While traditional methods, such as the finite element method, provide reliable solutions, their accuracy is often tied to the use of computationally intensive fine meshes. Moreover, they do not naturally account for measurements or prior solutions, and any change in the problem parameters requires results to be fully recomputed. They can also integrate prior knowledge and tackle entire families of PDEs by simply aggregating additional training losses. Nevertheless, they are highly sensitive to hyperparameters such as collocation points and the weights associated with each loss. It demonstrates that finding a (weak) solution of a PDE is equivalent to solving a constrained learning problem with worst-case losses. This explains the limitations of previous methods that minimize the expected value of aggregated losses. SCL also organically integrates structural constraints (e.g., invariances) and (partial) measurements or known solutions. The resulting constrained learning problems can be tackled using a practical algorithm that yields accurate solutions across a variety of PDEs, neural network architectures, and prior knowledge levels without extensive hyperparameter tuning and sometimes even at a lower computational cost. As such, a variety of numerical methods have been used to approximate their solutions, such as the celebrated finite element method (FEM). Despite their approximation guarantees, the accuracy of these methods is often in proportion to the computational complexity associated with finer discretizations. Additionally, they do not naturally incorporate prior knowledge, such as real-world measurements or previous solutions to similar equations. Lastly, any change to the PDE problem, such as initial condition or mesh size, requires the solution to be recomputed (Brenner & Scott, 2007; LeVeque, 2007; Katsikadelis, 2016). Methods based on neural network (NN) architectures, such as physics-informed NNs (PINNs) (Lagaris et al., 1998; Raissi et al., 2019; Lu et al., 2021b) and neural operators (NOs) (Li et al., 2021; Lu et al., 2021a; Rahman et al., 2023), have been proposed in an attempt to address these challenges. Rather than discretizing the PDE, these mesh-free approaches directly fit NNs to its solution.

The rise of large language models (LLMs) has led many researchers to consider their usage for scientific work. Some have found benefits using LLMs to augment or automate aspects of their research pipeline, while others have urged caution due to risks and ethical concerns. Yet little work has sought to quantify and characterize how researchers use LLMs and why. We present the first large-scale survey of 816 verified research article authors to understand how the research community leverages and perceives LLMs as research tools. We examine participants' self-reported LLM usage, finding that 81% of researchers have already incorporated LLMs into different aspects of their research workflow. We also find that traditionally disadvantaged groups in academia (non-White, junior, and non-native English speaking researchers) report higher LLM usage and perceived benefits, suggesting potential for improved research equity. However, women, non-binary, and senior researchers have greater ethical concerns, potentially hindering adoption.

Chain-of-thought (CoT) has proven to improve the reasoning capability of large language models (LLMs). However, due to the complexity of multimodal scenarios and the difficulty in collecting high-quality CoT data, CoT reasoning in multimodal LLMs has been largely overlooked. To this end, we propose a simple yet effective self-training framework, R3V, which iteratively enhances the model's Vision-language Reasoning by Reflecting on CoT Rationales. Our framework consists of two interleaved parts: (1) iteratively bootstrapping positive and negative solutions for reasoning datasets, and (2) reflection on rationale for learning from mistakes. Specifically, we introduce the self-refine and self-select losses, enabling the model to refine flawed rationale and derive the correct answer by comparing rationale candidates. Experiments on a wide range of vision-language tasks show that R3V consistently improves multimodal LLM reasoning, achieving a relative improvement of 23 to 60 percent over GPT-distilled baselines. Additionally, our approach supports self-reflection on generated solutions, further boosting performance through test-time computation.

Cognitive diagnosis (CD) utilizes students' existing studying records to estimate their mastery of unknown knowledge concepts, which is vital for evaluating their learning abilities. Accurate CD is extremely challenging because CD is associated with complex relationships and mechanisms among students, knowledge concepts, studying records, etc. However, existing approaches loosely consider these relationships and mechanisms by a non-end-to-end learning framework, resulting in sub-optimal feature extractions and fusions for CD. Different from them, this paper innovatively proposes an End-to-end Graph Neural Networks-based Cognitive Diagnosis (EGNN-CD) model. EGNN-CD consists of three main parts: knowledge concept network (KCN), graph neural networks-based feature extraction (GNNFE), and cognitive ability prediction (CAP). First, KCN constructs CD-related interaction by comprehensively extracting physical information from students, exercises, and knowledge concepts. Second, a four-channel GNNFE is designed to extract high-order and individual features from the constructed KCN. Finally, CAP employs a multi-layer perceptron to fuse the extracted features to predict students' learning abilities in an end-to-end learning way. With such designs, the feature extractions and fusions are guaranteed to be comprehensive and optimal for CD. Extensive experiments on three real datasets demonstrate that our EGNN-CD achieves significantly higher accuracy than state-of-the-art models in CD.

Using a graph to represent the team behavior ensures that the relationship between Multi-Armed Bandits (MABs) are a class of reinforcement the agents are held. However, existing works either do learning problems where an agent is presented with a set of not consider the weight of each relationship (graph edges) arms (i.e., actions), with each arm giving a reward drawn (Madhushani and Leonard, 2020; Agarwal et al., 2021) or from a probability distribution unknown to the agent expect the user to manually set those weights (Moradipari (Lattimore and Szepesvári, 2020). The goal of the agent et al., 2022). is to maximize its total reward which requires balancing In this paper, we propose a new approach that combines exploration and exploitation. MABs offer a simple model graph optimization and MAMAB algorithms to enhance to simulate decision-making under uncertainty. Practical team performance by expediting the convergence to consensus applications of MAB algorithms include news recommendations of arm means. Our proposed approach: (Yang and Toni, 2018), online ad placement (Aramayo et al., 2022), dynamic pricing (Babaioff et al., 2015), improves team performance by optimizing the edge and adaptive experimental design (Rafferty et al., 2019). In weights in the graph representing the team structure contrast to single-agent cases, in certain applications such in large constrained teams, as search and rescue, a team of agents should cooperate does not require manual tuning of the graph weights, with each other to accomplish goals by maximizing team is independent of the MAMAB algorithm and only performance. Such problems are solved using Multi-Agent depends on the consensus formula, and Multi-Armed Bandit (MAMAB) algorithms (Xu et al., formulates the problem as a convex optimization, which 2020). Most existing algorithms rely on the presence of is computationally efficient for large teams.

How can "weak teacher models" such as average human annotators or existing AI systems, effectively supervise LLMs to improve performance on hard reasoning tasks, especially those that challenge and requires expertise or daily practice from the teacher models? In this paper, we seek for empirical answers to this question by investigating various data-driven strategies that offer supervision data at different quality levels upon tasks of varying complexity. Two intuitive strategies emerge for teacher models to provide supervision during alignment training: 1) using lower-quality supervision from complete tasks that match the difficulty of the target reasoning tasks, and 2) leveraging higher-quality supervision from easier subtasks that are less challenging. Interestingly, we find that even when the outcome error rate for hard task supervision is high (e.g., 90\%), training on such data can outperform perfectly correct supervision on easier subtasks on multiple hard math benchmarks. We further identify a more critical factor influencing training performance: step-wise error rates, which indicate the severity of errors in solutions. Specifically, training on hard task supervision with the same outcome error rates but disparate step-wise error rates can lead to a 30\% accuracy gap on MATH benchmark. Our results also reveal that supplementing hard task supervision with the corresponding subtask supervision can yield notable performance improvements than simply combining rephrased hard full task supervision, suggesting new avenues for data augmentation. Data and code are released at \url{https://github.com/hexuan21/Weak-to-Strong}.

Large-scale deployment of large language models (LLMs) in various applications, such as chatbots and virtual assistants, requires LLMs to be culturally sensitive to the user to ensure inclusivity. Culture has been widely studied in psychology and anthropology, and there has been a recent surge in research on making LLMs more culturally inclusive in LLMs that goes beyond multilinguality and builds on findings from psychology and anthropology. In this paper, we survey efforts towards incorporating cultural awareness into text-based and multimodal LLMs. We start by defining cultural awareness in LLMs, taking the definitions of culture from anthropology and psychology as a point of departure. We then examine methodologies adopted for creating cross-cultural datasets, strategies for cultural inclusion in downstream tasks, and methodologies that have been used for benchmarking cultural awareness in LLMs. Further, we discuss the ethical implications of cultural alignment, the role of Human-Computer Interaction in driving cultural inclusion in LLMs, and the role of cultural alignment in driving social science research. We finally provide pointers to future research based on our findings about gaps in the literature.

Ontologies are useful for automatic machine processing of domain knowledge as they represent it in a structured format. Yet, constructing ontologies requires substantial manual effort. To automate part of this process, large language models (LLMs) have been applied to solve various subtasks of ontology learning. However, this partial ontology learning does not capture the interactions between subtasks. We address this gap by introducing OLLM, a general and scalable method for building the taxonomic backbone of an ontology from scratch. Rather than focusing on subtasks, like individual relations between entities, we model entire subcomponents of the target ontology by finetuning an LLM with a custom regulariser that reduces overfitting on high-frequency concepts. We introduce a novel suite of metrics for evaluating the quality of the generated ontology by measuring its semantic and structural similarity to the ground truth. In contrast to standard metrics, our metrics use deep learning techniques to define more robust distance measures between graphs. Both our quantitative and qualitative results on Wikipedia show that OLLM outperforms subtask composition methods, producing more semantically accurate ontologies while maintaining structural integrity. We further demonstrate that our model can be effectively adapted to new domains, like arXiv, needing only a small number of training examples. Our source code and datasets are available at https://github.com/andylolu2/ollm.

Agent learning from human interaction often relies on explicit signals, but implicit social cues, such as prosody in speech, could provide valuable information for more effective learning. This paper advocates for the integration of prosody as a teaching signal to enhance agent learning from human teachers. Through two exploratory studies--one examining voice feedback in an interactive reinforcement learning setup and the other analyzing restricted audio from human demonstrations in three Atari games--we demonstrate that prosody carries significant information about task dynamics. Our findings suggest that prosodic features, when coupled with explicit feedback, can enhance reinforcement learning outcomes. Moreover, we propose guidelines for prosody-sensitive algorithm design and discuss insights into teaching behavior. Our work underscores the potential of leveraging prosody as an implicit signal for more efficient agent learning, thus advancing human-agent interaction paradigms.