However, Spanish is occasionally adopted as the focus language for research endeavors and as result multiple projects are conducted in Spanish to explore language-specific nuances and challenges in NLP applications. Named-Entity recognition [1], Machine Translation [2], Semantic Relation Extraction [3] among others tasks have been conducted with a focus on Spanish language data, allowing for a more nuanced understanding of the intricacies involved. In this paper we present Classifier for Named Entities Recognized (CNER) a linguistically-aware online service that offers the possibility to test two main tasks of NLP, Named Entity Recognition (NER) and Relation Extraction (RE) for Spanish language. This together with other projects on Spanish language have been evaluated and adapted as a web service. In this context, language technologies and natural language processing (NLP) tools can support the identification of useful information in text and to promote its understanding. Specifically, CNER i) identifies the mentions follow the ACE standard with entity types include Person (PER), Organisation (ORG), Facility (FAC), Location (LOC), Geographical/Political (GPE), Vehicle (VEH), Vehicle (VEH) and Weapon (WEA) [4], [5]; ii) displays three different NER tools as previous step to RE task and iii) offers entity relationship information through tags GPE-AFF, PHYS, DISC, EMP-ORG, ART, NON-REL representing the relations between two entities [6] .

In this work, our goals are two fold: large-vocabulary continuous sign language recognition (CSLR), and sign language retrieval. To this end, we introduce a multi-task Transformer model, CSLR2, that is able to ingest a signing sequence and output in a joint embedding space between signed language and spoken language text. To enable CSLR evaluation in the large-vocabulary setting, we introduce new dataset annotations that have been manually collected. These provide continuous sign-level annotations for six hours of test videos, and will be made publicly available. We demonstrate that by a careful choice of loss functions, training the model for both the CSLR and retrieval tasks is mutually beneficial in terms of performance -- retrieval improves CSLR performance by providing context, while CSLR improves retrieval with more fine-grained supervision. We further show the benefits of leveraging weak and noisy supervision from large-vocabulary datasets such as BOBSL, namely sign-level pseudo-labels, and English subtitles. Our model significantly outperforms the previous state of the art on both tasks.

Understaffing has been consistently identified as the major challenge facing Healthcare today [7, 1, 2, 21, 55, 82, 97, 87, 124]. Automation tools that make use of Machine Learning (also known as Healthcare 4.0 [126]) have been consistently identified as crucial for reducing the workload of Healthcare professionals and improving the quality of care [5, 34, 44, 46, 78, 86, 94, 136]. In turn, the shortage of standard benchmarks has been consistently identified as a central roadblock for machine learning in Healthcare [27, 31, 49, 52, 59, 76, 81, 95, 110]. Whether it's ImageNet [32] in Computer Vision or GLUE [128] in natural language processing, benchmarks are a core research tool in mature applications of machine learning, enabling quantitative analysis of learning methodologies to guide and orient their development.

Federated learning (FL) requires edge devices to perform local training and exchange information with a parameter server, leading to substantial energy consumption. A critical challenge in practical FL systems is the rapid energy depletion of battery-limited edge devices, which curtails their operational lifespan and affects the learning performance. To address this issue, we apply energy harvesting technique in FL systems to extract ambient energy for continuously powering edge devices. We first establish the convergence bound for the wireless FL system with energy harvesting devices, illustrating that the convergence is impacted by partial device participation and packet drops, both of which depend on the energy supply. To accelerate the convergence, we formulate a joint device scheduling and power control problem and model it as a Markov decision process (MDP). By solving this MDP, we derive the optimal transmission policy and demonstrate that it possesses a monotone structure with respect to the battery and channel states. To overcome the curse of dimensionality caused by the exponential complexity of computing the optimal policy, we propose a low-complexity algorithm, which is asymptotically optimal as the number of devices increases. Furthermore, for unknown channels and harvested energy statistics, we develop a structure-enhanced deep reinforcement learning algorithm that leverages the monotone structure of the optimal policy to improve the training performance. Finally, extensive numerical experiments on real-world datasets are presented to validate the theoretical results and corroborate the effectiveness of the proposed algorithms.

Stochastic programming models can lead to very large-scale optimization problems for which it may be impossible to enumerate all possible scenarios. In such cases, one adopts a sampling-based solution methodology in which case the reliability of the resulting decisions may be suspect. For such instances, it is advisable to adopt methodologies that promote variance reduction. One such approach goes under a framework known as "compromise decision", which requires multiple replications of the solution procedure. This paper studies the reliability of stochastic programming solutions resulting from the "compromise decision" process. This process is characterized by minimizing an aggregation of objective function approximations across replications, presumably conducted in parallel. We refer to the post-parallel-processing problem as the problem of "compromise decision". We quantify the reliability of compromise decisions by estimating the expectation and variance of the "pessimistic distance" of sampled instances from the set of true optimal decisions. Such pessimistic distance is defined as an estimate of the largest possible distance of the solution of the sampled instance from the "true" optimal solution set. The Rademacher average of instances is used to bound the sample complexity of the compromise decision.

Chain of thought (CoT) has proven useful for problems requiring complex reasoning. Many of these problems are both textual and multimodal. Given the inputs in different modalities, a model generates a rationale and then uses it to answer a question. Because of the hallucination issue, the generated soft negative rationales with high textual quality but illogical semantics do not always help improve answer accuracy. This study proposes a rationale generation method using soft negative sampling (SNSE-CoT) to mitigate hallucinations in multimodal CoT. Five methods were applied to generate soft negative samples that shared highly similar text but had different semantics from the original. Bidirectional margin loss (BML) was applied to introduce them into the traditional contrastive learning framework that involves only positive and negative samples. Extensive experiments on the ScienceQA dataset demonstrated the effectiveness of the proposed method. Code and data are released at https://github.com/zgMin/SNSE-CoT.

Recent efforts have evaluated large language models (LLMs) in areas such as commonsense reasoning, mathematical reasoning, and code generation. However, to the best of our knowledge, no work has specifically investigated the performance of LLMs in natural language generation (NLG) tasks, a pivotal criterion for determining model excellence. Thus, this paper conducts a comprehensive evaluation of well-known and high-performing LLMs, namely ChatGPT, ChatGLM, T5-based models, LLaMA-based models, and Pythia-based models, in the context of NLG tasks. We select English and Chinese datasets encompassing Dialogue Generation and Text Summarization. Moreover, we propose a common evaluation setting that incorporates input templates and post-processing strategies. Our study reports both automatic results, accompanied by a detailed analysis.

Communication is defined as "Who says what to whom with what effect." A message from a communicator generates downstream receiver effects, also known as behavior. Receiver behavior, being a downstream effect of the message, carries rich signals about it. Even after carrying signals about the message, the behavior data is often ignored while training large language models. We show that training LLMs on receiver behavior can actually help improve their content-understanding abilities. Specifically, we show that training LLMs to predict the receiver behavior of likes and comments improves the LLM's performance on a wide variety of downstream content understanding tasks. We show this performance increase over 40 video and image understanding tasks over 23 benchmark datasets across both 0-shot and fine-tuning settings, outperforming many supervised baselines. Moreover, since receiver behavior, such as likes and comments, is collected by default on the internet and does not need any human annotations to be useful, the performance improvement we get after training on this data is essentially free-lunch. We release the receiver behavior cleaned comments and likes of 750k images and videos collected from multiple platforms along with our instruction-tuning data.

Concerns regarding Large Language Models (LLMs) to memorize and disclose private information, particularly Personally Identifiable Information (PII), become prominent within the community. Many efforts have been made to mitigate the privacy risks. However, the mechanism through which LLMs memorize PII remains poorly understood. To bridge this gap, we introduce a pioneering method for pinpointing PII-sensitive neurons (privacy neurons) within LLMs. Our method employs learnable binary weight masks to localize specific neurons that account for the memorization of PII in LLMs through adversarial training. Our investigations discover that PII is memorized by a small subset of neurons across all layers, which shows the property of PII specificity. Furthermore, we propose to validate the potential in PII risk mitigation by deactivating the localized privacy neurons. Both quantitative and qualitative experiments demonstrate the effectiveness of our neuron localization algorithm.

Accurate evaluation of financial question answering (QA) systems necessitates a comprehensive dataset encompassing diverse question types and contexts. However, current financial QA datasets lack scope diversity and question complexity. This work introduces FinTextQA, a novel dataset for long-form question answering (LFQA) in finance. FinTextQA comprises 1,262 high-quality, source-attributed QA pairs extracted and selected from finance textbooks and government agency websites.Moreover, we developed a Retrieval-Augmented Generation (RAG)-based LFQA system, comprising an embedder, retriever, reranker, and generator. A multi-faceted evaluation approach, including human ranking, automatic metrics, and GPT-4 scoring, was employed to benchmark the performance of different LFQA system configurations under heightened noisy conditions. The results indicate that: (1) Among all compared generators, Baichuan2-7B competes closely with GPT-3.5-turbo in accuracy score; (2) The most effective system configuration on our dataset involved setting the embedder, retriever, reranker, and generator as Ada2, Automated Merged Retrieval, Bge-Reranker-Base, and Baichuan2-7B, respectively; (3) models are less susceptible to noise after the length of contexts reaching a specific threshold.