We take the perspective in which we want to design a downstream task (such as estimating vegetation coverage) on a certain area of interest (AOI) with a limited labeling budget. By leveraging an existing Foundation Model (FM) we must decide whether we train a downstream model on a different but label-rich AOI hoping it generalizes to our AOI, or we split labels in our AOI for training and validating. In either case, we face choices concerning what FM to use, how to sample our AOI for labeling, etc. which affect both the performance and uncertainty of the results. In this work, we perform a large ablative study using eight existing FMs on either Sentinel 1 or Sentinel 2 as input data, and the classes from the ESA World Cover product as downstream tasks across eleven AOIs. We do repeated sampling and training, resulting in an ablation of some 500K simple linear regression models. Our results show both the limits of spatial generalizability across AOIs and the power of FMs where we are able to get over 0.9 correlation coefficient between predictions and targets on different chip level predictive tasks. And still, performance and uncertainty vary greatly across AOIs, tasks and FMs. We believe this is a key issue in practice, because there are many design decisions behind each FM and downstream task (input modalities, sampling, architectures, pretraining, etc.) and usually a downstream task designer is aware of and can decide upon a few of them. Through this work, we advocate for the usage of the methodology herein described (large ablations on reference global labels and simple probes), both when publishing new FMs, and to make informed decisions when designing downstream tasks to use them.

Small and medium-sized agricultural holders face challenges like limited access to localized, timely information, impacting productivity and sustainability. Traditional extension services, which rely on in-person agents, struggle with scalability and timely delivery, especially in remote areas. We introduce Farmer.Chat, a generative AI-powered chatbot designed to address these issues. Leveraging Generative AI, Farmer.Chat offers personalized, reliable, and contextually relevant advice, overcoming limitations of previous chatbots in deterministic dialogue flows, language support, and unstructured data processing. Deployed in four countries, Farmer.Chat has engaged over 15,000 farmers and answered over 300,000 queries. This paper highlights how Farmer.Chat's innovative use of GenAI enhances agricultural service scalability and effectiveness. Our evaluation, combining quantitative analysis and qualitative insights, highlights Farmer.Chat's effectiveness in improving farming practices, enhancing trust, response quality, and user engagement.

The complex nature of disease mechanisms and the variability of patient symptoms present significant obstacles in developing effective diagnostic tools. Although machine learning has made considerable advances in medical diagnosis, its decision-making processes frequently lack transparency, which can jeopardize patient outcomes. This underscores the critical need for Explainable AI (XAI), which not only offers greater clarity but also has the potential to significantly improve patient care. In this literature review, we conduct a detailed analysis of analyzing XAI methods identified through searches across various databases, focusing on chronic conditions such as Parkinson's, stroke, depression, cancer, heart disease, and Alzheimer's disease. The literature search revealed the application of 9 trending XAI algorithms in the field of healthcare and highlighted the pros and cons of each of them. Thus, the article is concluded with a critical appraisal of the challenges and future research opportunities for XAI in human health monitoring.

Selecting the best code solution from multiple generated ones is an essential task in code generation, which can be achieved by using some reliable validators (e.g., developer-written test cases) for assistance. Since reliable test cases are not always available and can be expensive to build in practice, researchers propose to automatically generate test cases to assess code solutions. However, when both code solutions and test cases are plausible and not reliable, selecting the best solution becomes challenging. Although some heuristic strategies have been proposed to tackle this problem, they lack a strong theoretical guarantee and it is still an open question whether an optimal selection strategy exists. Our work contributes in two ways. First, we show that within a Bayesian framework, the optimal selection strategy can be defined based on the posterior probability of the observed passing states between solutions and tests. The problem of identifying the best solution is then framed as an integer programming problem. Second, we propose an efficient approach for approximating this optimal (yet uncomputable) strategy, where the approximation error is bounded by the correctness of prior knowledge. We then incorporate effective prior knowledge to tailor code generation tasks. Both theoretical and empirical studies confirm that existing heuristics are limited in selecting the best solutions with plausible test cases. Our proposed approximated optimal strategy B4 significantly surpasses existing heuristics in selecting code solutions generated by large language models (LLMs) with LLM-generated tests, achieving a relative performance improvement by up to 50% over the strongest heuristic and 246% over the random selection in the most challenging scenarios. Our code is publicly available at https://github.com/ZJU-CTAG/B4.

Dynamical generative models that produce samples through an iterative process, such as Flow Matching and denoising diffusion models, have seen widespread use, but there has not been many theoretically-sound methods for improving these models with reward fine-tuning. In this work, we cast reward fine-tuning as stochastic optimal control (SOC). Critically, we prove that a very specific memoryless noise schedule must be enforced during fine-tuning, in order to account for the dependency between the noise variable and the generated samples. We also propose a new algorithm named Adjoint Matching which outperforms existing SOC algorithms, by casting SOC problems as a regression problem. We find that our approach significantly improves over existing methods for reward fine-tuning, achieving better consistency, realism, and generalization to unseen human preference reward models, while retaining sample diversity.

In recent years, there has been a surge in research on Question Difficulty Estimation (QDE) using natural language processing techniques. Transformer-based neural networks achieve state-of-the-art performance, primarily through supervised methods but with an isolated study in unsupervised learning. While supervised methods focus on predictive performance, they require abundant labeled data. On the other hand, unsupervised methods do not require labeled data but rely on a different evaluation metric that is also computationally expensive in practice. This work bridges the research gap by exploring active learning for QDE, a supervised human-in-the-loop approach striving to minimize the labeling efforts while matching the performance of state-of-the-art models. The active learning process iteratively trains on a labeled subset, acquiring labels from human experts only for the most informative unlabeled data points. Furthermore, we propose a novel acquisition function PowerVariance to add the most informative samples to the labeled set, a regression extension to the PowerBALD function popular in classification. We employ DistilBERT for QDE and identify informative samples by applying Monte Carlo dropout to capture epistemic uncertainty in unlabeled samples. The experiments demonstrate that active learning with PowerVariance acquisition achieves a performance close to fully supervised models after labeling only 10% of the training data. The proposed methodology promotes the responsible use of educational resources, makes QDE tools more accessible to course instructors, and is promising for other applications such as personalized support systems and question-answering tools.

Programs implemented in various programming languages form the foundation of software applications. To alleviate the burden of program migration and facilitate the development of software systems, automated program translation across languages has garnered significant attention. Previous approaches primarily focus on pairwise translation paradigms, learning translation between pairs of languages using bilingual parallel data. However, parallel data is difficult to collect for some language pairs, and the distribution of program semantics across languages can shift, posing challenges for pairwise program translation. In this paper, we argue that jointly learning a unified model to translate code across multiple programming languages is superior to separately learning from bilingual parallel data. We propose Variational Interaction for Multilingual Program Translation~(VIM-PT), a disentanglement-based generative approach that jointly trains a unified model for multilingual program translation across multiple languages. VIM-PT disentangles code into language-shared and language-specific features, using variational inference and interaction information with a novel lower bound, then achieves program translation through conditional generation. VIM-PT demonstrates four advantages: 1) captures language-shared information more accurately from various implementations and improves the quality of multilingual program translation, 2) mines and leverages the capability of non-parallel data, 3) addresses the distribution shift of program semantics across languages, 4) and serves as a unified model, reducing deployment complexity.

Recent advancements in large language models (LLMs) have showcased impressive code generation capabilities, primarily evaluated through language-to-code benchmarks. However, these benchmarks may not fully capture a model's code understanding abilities. We introduce CodeJudge-Eval (CJ-Eval), a novel benchmark designed to assess LLMs' code understanding abilities from the perspective of code judging rather than code generation. CJ-Eval challenges models to determine the correctness of provided code solutions, encompassing various error types and compilation issues. By leveraging a diverse set of problems and a fine-grained judging system, CJ-Eval addresses the limitations of traditional benchmarks, including the potential memorization of solutions. Evaluation of 12 well-known LLMs on CJ-Eval reveals that even state-of-the-art models struggle, highlighting the benchmark's ability to probe deeper into models' code understanding abilities. Our codes and benchmark are available at \url{https://github.com/CodeLLM-Research/CodeJudge-Eval}.

While large language models (LLMs) have shown promise for medical question answering, there is limited work focused on tropical and infectious disease-specific exploration. We build on an opensource tropical and infectious diseases (TRINDs) dataset, expanding it to include demographic and semantic clinical and consumer augmentations yielding 11000+ prompts. We evaluate LLM performance on these, comparing generalist and medical LLMs, as well as LLM outcomes to human experts. We demonstrate through systematic experimentation, the benefit of contextual information such as demographics, location, gender, risk factors for optimal LLM response. Finally we develop a prototype of TRINDs-LM, a research tool that provides a playground to navigate how context impacts LLM outputs for health.

It is unquestionable that time series forecasting is of paramount importance in many fields. The most used machine learning models to address time series forecasting tasks are Recurrent Neural Networks (RNNs). Typically, those models are built using one of the three most popular cells: ELMAN, Long Short-Term Memory (LSTM), or Gated Recurrent Unit (GRU) cells. Each cell has a different structure and implies a different computational cost. However, it is not clear why and when to use each RNN-cell structure. Actually, there is no comprehensive characterization of all the possible time series behaviors and no guidance on what RNN cell structure is the most suitable for each behavior. The objective of this study is twofold: it presents a comprehensive taxonomy of almost all time series behaviors and provides insights into the best RNN cell structure for each time series behavior. We conducted two experiments: (1) We evaluate and analyze the role of each component in the LSTM-Vanilla cell by creating 11 variants based on one alteration in its basic architecture (removing, adding, or substituting one cell component). (2) We evaluate and analyze the performance of 20 possible RNN-cell structures. To evaluate, compare, and select the best model, different statistical metrics were used: error-based metrics, information criterion-based metrics, naive-based metrics, and direction change-based metrics. To further improve our confidence in the models interpretation and selection, the Friedman Wilcoxon-Holm signed-rank test was used. Our results advocate the usage and exploration of the newly created RNN variant, named SLIM, in time series forecasting thanks to its high ability to accurately predict the different time series behaviors, as well as its simple structural design that does not require expensive temporal and computing resources.