Classic Delphi and Fuzzy Delphi methods are used to test content validity of a data collection tools such as questionnaires. Fuzzy Delphi takes the opinion issued by judges from a linguistic perspective reducing ambiguity in opinions by using fuzzy numbers. We propose an extension named 2-Tuple Fuzzy Linguistic Delphi method to deal with scenarios in which judges show different expertise degrees by using fuzzy multigranular semantics of the linguistic terms and to obtain intermediate and final results expressed by 2-tuple linguistic values. The key idea of our proposal is to validate the full questionnaire by means of the evaluation of its parts, defining the validity of each item as a Decision Making problem. Taking the opinion of experts, we measure the degree of consensus, the degree of consistency, and the linguistic score of each item, in order to detect those items that affect, positively or negatively, the quality of the instrument. Considering the real need to evaluate a b-learning educational experience with a consensual questionnaire, we present a Decision Making model for questionnaire validation that solve it. Additionally, we contribute to this consensus reaching problem by developing an online tool under GPL v3 license. The software visualizes the collective valuations for each iteration and assists to determine which parts of the questionnaire should be modified to reach a consensual solution.

This work presents the early development of a model of image captioning for the Brazilian Portuguese language. We used the GRIT (Grid - and Region-based Image captioning Transformer) model to accomplish this work. GRIT is a Transformer-only neural architecture that effectively utilizes two visual features to generate better captions. The GRIT method emerged as a proposal to be a more efficient way to generate image captioning. In this work, we adapt the GRIT model to be trained in a Brazilian Portuguese dataset to have an image captioning method for the Brazilian Portuguese Language.

We apply a state-of-the-art membership inference attack (MIA) to systematically test the practical privacy vulnerability of fine-tuning large image classification models.We focus on understanding the properties of data sets and samples that make them vulnerable to membership inference. In terms of data set properties, we find a strong power law dependence between the number of examples per class in the data and the MIA vulnerability, as measured by true positive rate of the attack at a low false positive rate. For an individual sample, large gradients at the end of training are strongly correlated with MIA vulnerability.

An open scientific challenge is how to classify events with reliable measures of uncertainty, when we have a mechanistic model of the data-generating process but the distribution over both labels and latent nuisance parameters is different between train and target data. We refer to this type of distributional shift as generalized label shift (GLS). Direct classification using observed data $\mathbf{X}$ as covariates leads to biased predictions and invalid uncertainty estimates of labels $Y$. We overcome these biases by proposing a new method for robust uncertainty quantification that casts classification as a hypothesis testing problem under nuisance parameters. The key idea is to estimate the classifier's receiver operating characteristic (ROC) across the entire nuisance parameter space, which allows us to devise cutoffs that are invariant under GLS. Our method effectively endows a pre-trained classifier with domain adaptation capabilities and returns valid prediction sets while maintaining high power. We demonstrate its performance on two challenging scientific problems in biology and astroparticle physics with data from realistic mechanistic models.

Symbolic regression holds great potential for uncovering underlying mathematical and physical relationships from empirical data. While existing transformer-based models have recently achieved significant success in this domain, they face challenges in terms of generalizability and adaptability. Typically, in cases where the output expressions do not adequately fit experimental data, the models lack efficient mechanisms to adapt or modify the expression. This inflexibility hinders their application in real-world scenarios, particularly in discovering unknown physical or biological relationships. Inspired by how human experts refine and adapt expressions, we introduce Symbolic Q-network (Sym-Q), a novel reinforcement learning-based model that redefines symbolic regression as a sequential decision-making task. Sym-Q leverages supervised demonstrations and refines expressions based on reward signals indicating the quality of fitting precision. Its distinctive ability to manage the complexity of expression trees and perform precise step-wise updates significantly enhances flexibility and efficiency. Our results demonstrate that Sym-Q excels not only in recovering underlying mathematical structures but also uniquely learns to efficiently refine the output expression based on reward signals, thereby discovering underlying expressions. Sym-Q paves the way for more intuitive and impactful discoveries in physical science, marking a substantial advancement in the field of symbolic regression.

Accurate brain tumor segmentation in the early stages of the disease is crucial for the treatment's effectiveness, avoiding exhaustive visual inspection of a qualified specialist on 3D MR brain images of multiple protocols (e.g., T1, T2, T2-FLAIR, T1-Gd). Several networks exist for Glioma segmentation, being nnU-Net one of the best. In this work, we evaluate self-calibrated convolutions in different parts of the nnU-Net network to demonstrate that self-calibrated modules in skip connections can significantly improve the enhanced-tumor and tumor-core segmentation accuracy while preserving the wholetumor segmentation accuracy.

Large Language Models (LLMs) create exciting possibilities for powerful language processing tools to accelerate research in materials science. While LLMs have great potential to accelerate materials understanding and discovery, they currently fall short in being practical materials science tools. In this position paper, we show relevant failure cases of LLMs in materials science that reveal current limitations of LLMs related to comprehending and reasoning over complex, interconnected materials science knowledge. Given those shortcomings, we outline a framework for developing Materials Science LLMs (MatSci-LLMs) that are grounded in materials science knowledge and hypothesis generation followed by hypothesis testing. The path to attaining performant MatSci-LLMs rests in large part on building high-quality, multi-modal datasets sourced from scientific literature where various information extraction challenges persist. As such, we describe key materials science information extraction challenges which need to be overcome in order to build large-scale, multi-modal datasets that capture valuable materials science knowledge. Finally, we outline a roadmap for applying future MatSci-LLMs for real-world materials discovery via: 1. Automated Knowledge Base Generation; 2. Automated In-Silico Material Design; and 3. MatSci-LLM Integrated Self-Driving Materials Laboratories.

Artificial intelligence (AI) has driven many information and communication technology (ICT) breakthroughs. Nonetheless, the scope of ICT systems has expanded far beyond AI since the Turing test proposal. Critically, recent AI regulation proposals adopt AI definitions affecting ICT techniques, approaches, and systems that are not AI. In some cases, even works from mathematics, statistics, and engineering would be affected. Worryingly, AI misdefinitions are observed from Western societies to the Global South. In this paper, we propose a framework to score how \textit{validated as appropriately-defined for regulation} (VADER) an AI definition is. Our online, publicly-available VADER framework scores the coverage of premises that should underlie AI definitions for regulation, which aim to (i) reproduce principles observed in other successful technology regulations, and (ii) include all AI techniques and approaches while excluding non-AI works. Regarding the latter, our score is based on a dataset of representative AI, non-AI ICT, and non-ICT examples. We demonstrate our contribution by reviewing the AI regulation proposals of key players, namely the United States, United Kingdom, European Union, and Brazil. Importantly, none of the proposals assessed achieve the appropriateness score, ranging from a revision need to a concrete risk to ICT systems and works from other fields.

Cognitive assistants (CA) are chatbots that provide context-aware support to human workers in knowledge-intensive tasks. Traditionally, cognitive assistants respond in specific ways to predefined user intents and conversation patterns. However, this rigidness does not handle the diversity of natural language well. Recent advances in natural language processing (NLP), powering large language models (LLM) such as GPT-4, Llama2, and Gemini, could enable CAs to converse in a more flexible, human-like manner. However, the additional degrees of freedom may have unforeseen consequences, especially in knowledge-intensive contexts where accuracy is crucial. As a preliminary step to assessing the potential of using LLMs in these contexts, we conducted a user study comparing an LLM-based CA to an intent-based system regarding interaction efficiency, user experience, workload, and usability. This revealed that LLM-based CAs exhibited better user experience, task completion rate, usability, and perceived performance than intent-based systems, suggesting that switching NLP techniques should be investigated further.

There is a consensus that instruction fine-tuning of LLMs requires high-quality data, but what are they? LIMA (NeurIPS 2023) and AlpaGasus (ICLR 2024) are state-of-the-art methods for selecting such high-quality examples, either via manual curation or using GPT-3.5-Turbo as a quality scorer. We show that the extremely simple baseline of selecting the 1,000 instructions with longest responses from standard datasets can consistently outperform these sophisticated methods according to GPT-4 and PaLM-2 as judges, while remaining competitive on the OpenLLM benchmarks that test factual knowledge. We demonstrate this for several state-of-the-art LLMs (Llama-2-7B, Llama-2-13B, and Mistral-7B) and datasets (Alpaca-52k and Evol-Instruct-70k). In addition, a lightweight refinement of such long instructions can further improve the abilities of the fine-tuned LLMs, and allows us to obtain the 2nd highest-ranked Llama-2-7B-based model on AlpacaEval 2.0 while training on only 1,000 examples and no extra preference data. We also conduct a thorough analysis of our models to ensure that their enhanced performance is not simply due to GPT-4's preference for longer responses, thus ruling out any artificial improvement. In conclusion, our findings suggest that fine-tuning on the longest instructions should be the default baseline for any research on instruction fine-tuning.