Progress in scientific discovery is rarely the result of a single "Eureka" moment, but is rather the product of hundreds of scientists incrementally working together toward a common goal. While existing agent workflows are capable of producing research autonomously, they do so in isolation, without the ability to continuously improve upon prior research results. To address these challenges, we introduce AgentRxiv-a framework that lets LLM agent laboratories upload and retrieve reports from a shared preprint server in order to collaborate, share insights, and iteratively build on each other's research. We task agent laboratories to develop new reasoning and prompting techniques and find that agents with access to their prior research achieve higher performance improvements compared to agents operating in isolation (11.4% relative improvement over baseline on MATH-500). We find that the best performing strategy generalizes to benchmarks in other domains (improving on average by 3.3%). Multiple agent laboratories sharing research through AgentRxiv are able to work together towards a common goal, progressing more rapidly than isolated laboratories, achieving higher overall accuracy (13.7% relative improvement over baseline on MATH-500). These findings suggest that autonomous agents may play a role in designing future AI systems alongside humans. We hope that AgentRxiv allows agents to collaborate toward research goals and enables researchers to accelerate discovery.

Deep neural networks have been shown to learn and rely on spurious correlations present in the data that they are trained on. Reliance on such correlations can cause these networks to malfunction when deployed in the real world, where these correlations may no longer hold. To overcome the learning of and reliance on such correlations, recent studies propose approaches that yield promising results. These works, however, study settings where the strength of the spurious signal is significantly greater than that of the core, invariant signal, making it easier to detect the presence of spurious features in individual training samples and allow for further processing. In this paper, we identify new settings where the strength of the spurious signal is relatively weaker, making it difficult to detect any spurious information while continuing to have catastrophic consequences. We also discover that spurious correlations are learned primarily due to only a handful of all the samples containing the spurious feature and develop a novel data pruning technique that identifies and prunes small subsets of the training data that contain these samples. Our proposed technique does not require inferred domain knowledge, information regarding the sample-wise presence or nature of spurious information, or human intervention. Finally, we show that such data pruning attains state-of-the-art performance on previously studied settings where spurious information is identifiable.

Large Language Models (LLMs) have demonstrated exceptional performance on a range of downstream NLP tasks by generating text that closely resembles human writing. However, the ease of achieving this similarity raises concerns from potential malicious uses at scale by bad actors, as LLM-generated text becomes increasingly difficult to discern from human text. Although detection methods have been developed to address this issue, bad actors can further manipulate LLM-generated texts to make them less detectable. In this work, we study how further editing texts with Reinforcement Learning from Human Feedback (RLHF), which aligns model outputs with human preferences, affects (a) the quality of generated texts for two tasks, and (b) the performance of LLM-generated text detectors, looking at both training-based and zero-shot detection methods. Although RLHF improves the quality of LLM-generated texts, we find that it also tends to produce more detectable, lengthy, and repetitive outputs. Additionally, we observe that training-based detectors are vulnerable to short texts and to texts that incorporate code, whereas zero-shot detectors exhibit greater robustness.

Background Advances in machine learning (ML) models have increased the capability of researchers to detect vaccine hesitancy in social media using Natural Language Processing (NLP). A considerable volume of research has identified the persistence of COVID-19 vaccine hesitancy in discourse shared on various social media platforms. Methods Our objective in this study was to conduct a systematic review of research employing sentiment analysis or stance detection to study discourse towards COVID-19 vaccines and vaccination spread on Twitter (officially known as X since 2023). Following registration in the PROSPERO international registry of systematic reviews, we searched papers published from 1 January 2020 to 31 December 2023 that used supervised machine learning to assess COVID-19 vaccine hesitancy through stance detection or sentiment analysis on Twitter. We categorized the studies according to a taxonomy of five dimensions: tweet sample selection approach, self-reported study type, classification typology, annotation codebook definitions, and interpretation of results. We analyzed if studies using stance detection report different hesitancy trends than those using sentiment analysis by examining how COVID-19 vaccine hesitancy is measured, and whether efforts were made to avoid measurement bias. Results Our review found that measurement bias is widely prevalent in studies employing supervised machine learning to analyze sentiment and stance toward COVID-19 vaccines and vaccination. The reporting errors are sufficiently serious that they hinder the generalisability and interpretation of these studies to understanding whether individual opinions communicate reluctance to vaccinate against SARS-CoV-2. Conclusion Improving the reporting of NLP methods is crucial to addressing knowledge gaps in vaccine hesitancy discourse.

Large Language Models (LLMs) have significantly advanced various fields, particularly coding, mathematical reasoning, and logical problem solving. However, a critical question remains: Do these mathematical reasoning abilities persist when LLMs are presented with culturally adapted math problems? Specifically, how do LLMs perform when faced with math problems embedded in cultural contexts that have no significant representation in main stream web-scale AI training data? To explore this, we generated six synthetic cultural datasets from GSM8K, a widely used benchmark for assessing LLMs' mathematical reasoning skills. While preserving the mathematical logic and numerical values of the original GSM8K test set, we modify cultural elements such as personal names, food items, place names, etc. These culturally adapted datasets provide a more reliable framework for evaluating LLMs' mathematical reasoning under shifting cultural contexts. Our findings reveal that LLMs struggle with math problems when cultural references change, even though the underlying mathematical structure remains constant. Smaller models exhibit greater performance drops compared to larger models. Interestingly, our results also suggest that cultural familiarity can enhance mathematical reasoning. Even models with no explicit mathematical training but exposure to relevant cultural contexts sometimes outperform larger, mathematically proficient models on culturally embedded math problems. This study highlights the impact of cultural context on the mathematical reasoning abilities of LLMs, underscoring the need for more diverse and representative training data to improve robustness in real-world applications. The benchmark data sets and script for reproducing the results are available at https://github.com/akarim23131/Lost_in_Cultural_Translation

Multi-task prompt tuning utilizes multiple high-resource source tasks to improve performance on low-source target tasks. Existing approaches transfer the soft prompt trained by combining all source tasks or a single ``high-similar'' source task one-time-only. However, we find that the optimal transfer performance often comes from a combination of source tasks, which is neither one nor all. Further, we find that the similarity between source and target tasks also changes dynamically during fine-tuning after transfering, making similarity calculation in the initiation stage inadequate. To address these issues, we propose a method called Dynamic Task Vector Grouping (DTVG), whose core ideas contain (1) measuring the task similarity with task vectors instead of soft prompt, (2) grouping the optimal source task combination based on two metrics: {\it target similarity} and {\it knowledge consistency}; (3) dynamically updating the combination in each iteration step. Extensive experiments on the 26 NLP datasets under different settings demonstrate that DTVG effectively groups similar source tasks while reducing negative transfer, achieving the start-of-art performance.

Electrochemical interfaces are crucial in catalysis, energy storage, and corrosion, where their stability and reactivity depend on complex interactions between the electrode, adsorbates, and electrolyte. Predicting stable surface structures remains challenging, as traditional surface Pourbaix diagrams tend to either rely on expert knowledge or costly $\textit{ab initio}$ sampling, and neglect thermodynamic equilibration with the environment. Machine learning (ML) potentials can accelerate static modeling but often overlook dynamic surface transformations. Here, we extend the Virtual Surface Site Relaxation-Monte Carlo (VSSR-MC) method to autonomously sample surface reconstructions modeled under aqueous electrochemical conditions. Through fine-tuning foundational ML force fields, we accurately and efficiently predict surface energetics, recovering known Pt(111) phases and revealing new LaMnO$_\mathrm{3}$(001) surface reconstructions. By explicitly accounting for bulk-electrolyte equilibria, our framework enhances electrochemical stability predictions, offering a scalable approach to understanding and designing materials for electrochemical applications.

In this research, renewable energy expansion in South America up to 2050 is predicted based on machine learning models that are trained on past energy data. The research employs gradient boosting regression and Prophet time series forecasting to make predictions of future generation capacities for solar, wind, hydroelectric, geothermal, biomass, and other renewable sources in South American nations. Model output analysis indicates staggering future expansion in the generation of renewable energy, with solar and wind energy registering the highest expansion rates. Geospatial visualization methods were applied to illustrate regional disparities in the utilization of renewable energy. The results forecast South America to record nearly 3-fold growth in the generation of renewable energy by the year 2050, with Brazil and Chile spearheading regional development. Such projections help design energy policy, investment strategy, and climate change mitigation throughout the region, in helping the developing economies to transition to sustainable energy.

Road fatalities pose significant public safety and health challenges worldwide, with pedestrians being particularly vulnerable in vehicle-pedestrian crashes due to disparities in physical and performance characteristics. This study employs explainable artificial intelligence (XAI) to identify key factors contributing to pedestrian fatalities across the five U.S. states with the highest crash rates (2018-2022). It compares them to the five states with the lowest fatality rates. Using data from the Fatality Analysis Reporting System (FARS), the study applies machine learning techniques-including Decision Trees, Gradient Boosting Trees, Random Forests, and XGBoost-to predict contributing factors to pedestrian fatalities. To address data imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) is utilized, while SHapley Additive Explanations (SHAP) values enhance model interpretability. The results indicate that age, alcohol and drug use, location, and environmental conditions are significant predictors of pedestrian fatalities. The XGBoost model outperformed others, achieving a balanced accuracy of 98 %, accuracy of 90 %, precision of 92 %, recall of 90 %, and an F1 score of 91 %. Findings reveal that pedestrian fatalities are more common in mid-block locations and areas with poor visibility, with older adults and substance-impaired individuals at higher risk. These insights can inform policymakers and urban planners in implementing targeted safety measures, such as improved lighting, enhanced pedestrian infrastructure, and stricter traffic law enforcement, to reduce fatalities and improve public safety.

Many high-impact machine learning tasks involve multi-dimensional data (e.g., images, volumetric medical scans, multivariate time-series). Yet, most neural architectures flatten inputs, discarding critical cross-dimension information. We introduce NdLinear, a novel linear transformation that preserves these structures without extra overhead. By operating separately along each dimension, NdLinear captures dependencies that standard fully connected layers overlook. Extensive experiments across convolutional, recurrent, and transformer-based networks show significant improvements in representational power and parameter efficiency. Crucially, NdLinear serves as a foundational building block for large-scale foundation models by operating on any unimodal or multimodal data in its native form. This removes the need for flattening or modality-specific preprocessing. Ndlinear rethinks core architectural priorities beyond attention, enabling more expressive, context-aware models at scale. We propose NdLinear as a drop-in replacement for standard linear layers -- marking an important step toward next-generation neural architectures.