The Fon language, spoken by an average 2 million of people, is a truly low-resourced African language, with a limited online presence, and existing datasets (just to name but a few). Multitask learning is a learning paradigm that aims to improve the generalization capacity of a model by sharing knowledge across different but related tasks: this could be prevalent in very data-scarce scenarios. In this paper, we present the first explorative approach to multitask learning, for model capabilities enhancement in Natural Language Processing for the Fon language. Specifically, we explore the tasks of Named Entity Recognition (NER) and Part of Speech Tagging (POS) for Fon. We leverage two language model heads as encoders to build shared representations for the inputs, and we use linear layers blocks for classification relative to each task. Our results on the NER and POS tasks for Fon, show competitive (or better) performances compared to several multilingual pretrained language models finetuned on single tasks. Additionally, we perform a few ablation studies to leverage the efficiency of two different loss combination strategies and find out that the equal loss weighting approach works best in our case. Our code is open-sourced at https://github.com/bonaventuredossou/multitask_fon.

Reinforcement learning from human feedback (RLHF) is a technique for training AI systems to align with human goals. RLHF has emerged as the central method used to finetune state-of-the-art large language models (LLMs). Despite this popularity, there has been relatively little public work systematizing its flaws. In this paper, we (1) survey open problems and fundamental limitations of RLHF and related methods; (2) overview techniques to understand, improve, and complement RLHF in practice; and (3) propose auditing and disclosure standards to improve societal oversight of RLHF systems. Our work emphasizes the limitations of RLHF and highlights the importance of a multi-faceted approach to the development of safer AI systems.

ABSTRACT Computational models help decision makers understand epidemic dynamics to optimize public health interventions. Agent-based simulation of disease spread in synthetic populations allows us to compare and contrast different effects across identical populations or to investigate the effect of interventions keeping every other factor constant between "digital twins". In particular, we can observe the behavior of two different diseases as they evolve from identical initial conditions in the same population. FRED (A Framework for Reconstructing Epidemiological Dynamics) is an agent-based modeling system with a geospatial perspective using a synthetic population that is constructed based on the U.S. census data. Having synthetic data provides a baseline to get comparable results from different conditions and interventions. In this paper we show how Gaussian process regression can be used on FRED-synthesized data to infer the differing spatial dispersion of the epidemic dynamics for two disease conditions that start from the same initial conditions and spread among identical populations. Our results showcase the utility of agent-based simulations frameworks such as FRED for inferring differences between conditions where controlling for all confounding factors for such comparisons is next to impossible without synthetic data.

Simulating and predicting water levels in river systems is essential for flood warnings, hydraulic operations, and flood mitigations. In the engineering field, tools such as HEC-RAS, MIKE, and SWMM are used to build detailed physics-based hydrological and hydraulic computational models to simulate the entire watershed, thereby predicting the water stage at any point in the system. However, these physics-based models are computationally intensive, especially for large watersheds and for longer simulations. To overcome this problem, we train several deep learning (DL) models for use as surrogate models to rapidly predict the water stage. The downstream stage of the Miami River in South Florida is chosen as a case study for this paper. The dataset is from January 1, 2010, to December 31, 2020, downloaded from the DBHYDRO database of the South Florida Water Management District (SFWMD). Extensive experiments show that the performance of the DL models is comparable to that of the physics-based models, even during extreme precipitation conditions (i.e., tropical storms). Furthermore, we study the decline in prediction accuracy of the DL models with an increase in prediction lengths. In order to predict the water stage in the future, our DL models use measured variables of the river system from the recent past as well as covariates that can be reliably predicted in the near future. In summary, the deep learning models achieve comparable or better error rates with at least 1000x speedup in comparison to the physics-based models.

In this paper, we present a transformer architecture for predicting student performance on standardized tests. Specifically, we leverage students historical data, including their past test scores, study habits, and other relevant information, to create a personalized model for each student. We then use these models to predict their future performance on a given test. Applying this model to the RIIID dataset, we demonstrate that using multiple granularities for temporal features as the decoder input significantly improve model performance. Our results also show the effectiveness of our approach, with substantial improvements over the LightGBM method. Our work contributes to the growing field of AI in education, providing a scalable and accurate tool for predicting student outcomes.

Real-world time-series datasets are often multivariate with complex dynamics. To capture this complexity, high capacity architectures like recurrent- or attention-based sequential deep learning models have become popular. However, recent work demonstrates that simple univariate linear models can outperform such deep learning models on several commonly used academic benchmarks. Extending them, in this paper, we investigate the capabilities of linear models for time-series forecasting and present Time-Series Mixer (TSMixer), a novel architecture designed by stacking multi-layer perceptrons (MLPs). TSMixer is based on mixing operations along both the time and feature dimensions to extract information efficiently. On popular academic benchmarks, the simple-to-implement TSMixer is comparable to specialized state-of-the-art models that leverage the inductive biases of specific benchmarks. On the challenging and large scale M5 benchmark, a real-world retail dataset, TSMixer demonstrates superior performance compared to the state-of-the-art alternatives. Our results underline the importance of efficiently utilizing cross-variate and auxiliary information for improving the performance of time series forecasting. We present various analyses to shed light into the capabilities of TSMixer. The design paradigms utilized in TSMixer are expected to open new horizons for deep learning-based time series forecasting. The implementation is available at https://github.com/google-research/google-research/tree/master/tsmixer

Non-differentiable controllers and rule-based policies are widely used for controlling real systems such as telecommunication networks and robots. Specifically, parameters of mobile network base station antennas can be dynamically configured by these policies to improve users coverage and quality of service. Motivated by the antenna tilt control problem, we introduce Model-Based Residual Policy Learning (MBRPL), a practical reinforcement learning (RL) method. MBRPL enhances existing policies through a model-based approach, leading to improved sample efficiency and a decreased number of interactions with the actual environment when compared to off-the-shelf RL methods.To the best of our knowledge, this is the first paper that examines a model-based approach for antenna control. Experimental results reveal that our method delivers strong initial performance while improving sample efficiency over previous RL methods, which is one step towards deploying these algorithms in real networks.

Studying data memorization in neural language models helps us understand the risks (e.g., to privacy or copyright) associated with models regurgitating training data and aids in the development of countermeasures. Many prior works -- and some recently deployed defenses -- focus on "verbatim memorization", defined as a model generation that exactly matches a substring from the training set. We argue that verbatim memorization definitions are too restrictive and fail to capture more subtle forms of memorization. Specifically, we design and implement an efficient defense that perfectly prevents all verbatim memorization. And yet, we demonstrate that this "perfect" filter does not prevent the leakage of training data. Indeed, it is easily circumvented by plausible and minimally modified "style-transfer" prompts -- and in some cases even the non-modified original prompts -- to extract memorized information. We conclude by discussing potential alternative definitions and why defining memorization is a difficult yet crucial open question for neural language models.

Count data play a critical role in medical research, such as heart disease. The Poisson regression model is a common technique for evaluating the impact of a set of covariates on the count responses. The mixture of Poisson regression models with experts is a practical tool to exploit the covariates, not only to handle the heterogeneity in the Poisson regressions but also to learn the mixing structure of the population. Multicollinearity is one of the most common challenges with regression models, leading to ill-conditioned design matrices of Poisson regression components and expert classes. The maximum likelihood method produces unreliable and misleading estimates for the effects of the covariates in multicollinearity. In this research, we develop Ridge and Liu-type methods as two shrinkage approaches to cope with the ill-conditioned design matrices of the mixture of Poisson regression models with experts. Through various numerical studies, we demonstrate that the shrinkage methods offer more reliable estimates for the coefficients of the mixture model in multicollinearity while maintaining the classification performance of the ML method. The shrinkage methods are finally applied to a heart study to analyze the heart disease rate stages.

A drone attack on an open market south of the Sudanese capital, Khartoum, has killed at least 40 people, activists and medical workers said, as the military and the paramilitary Rapid Support Forces (RSF) battle for control of the country. At least 70 others were injured in the attack in Khartoum's Mayo neighbourhood on Sunday, according to resistance committees and two healthcare workers at the Bashair University Hospital, where the casualties were treated. Many of them will require amputations. The group posted footage on social media showing bodies wrapped in white sheets in an open yard at the hospital. Reporting from Khartoum, Al Jazeera's Hiba Morgan said the drone attack was carried out by the Sudanese army.