Recently, continuous diffusion models (CDM) have been introduced into non-autoregressive (NAR) text-to-text generation. However, the discrete nature of text increases the difficulty of CDM to generate coherent and fluent texts, and also causes the incompatibility problem between CDM and advanced NLP techniques, especially the popular pre-trained language models~(PLMs). To solve it, we propose Diffusion-NAT, which introduces discrete diffusion models~(DDM) into NAR text-to-text generation and integrates BART to improve the performance. By revising the decoding process of BART and the typical settings of DDM, we unify the inference process of BART and the denoising process of DDM into the same NAR masked tokens recovering task. In this way, DDM can rely on BART to perform denoising, which can benefit from both the rich pre-learned knowledge of BART and the iterative refining paradigm of DDM. Besides, we also propose the iterative self-prompting strategy to further improve the generation quality. Experimental results on 7 datasets show that our approach can outperform competitive NAR methods, and even surpass autoregressive methods. Our code and data will be publicly released.

Prompt tuning is a parameter-efficient approach to adapting pre-trained language models to downstream tasks. Although prompt tuning has been shown to match the performance of full model tuning when training data is sufficient, it tends to struggle in few-shot learning settings. In this paper, we present Multi-task Pre-trained Modular Prompt (MP2) to boost prompt tuning for few-shot learning. MP2 is a set of combinable prompts pre-trained on 38 Chinese tasks. On downstream tasks, the pre-trained prompts are selectively activated and combined, leading to strong compositional generalization to unseen tasks. To bridge the gap between pre-training and fine-tuning, we formulate upstream and downstream tasks into a unified machine reading comprehension task. Extensive experiments under two learning paradigms, i.e., gradient descent and black-box tuning, show that MP2 significantly outperforms prompt tuning, full model tuning, and prior prompt pre-training methods in few-shot settings. In addition, we demonstrate that MP2 can achieve surprisingly fast and strong adaptation to downstream tasks by merely learning 8 parameters to combine the pre-trained modular prompts.

PETA has given the Bible's Book of Genesis a "vegan" makeover, using ChatGPT to recreate the story and "send a can't-be-missed animal rights message filled with vegan teachings." In PETA's vegan version of the Bible, animals are referred to as "beings" rather than "beasts" or "creatures" and plants like hemp and bamboo are used for clothing instead of animal skins because "no one with any fashion or moral sense would wear animal skins in the 21st century." Volunteers serve vegan hot dogs at the PETA Congressional Veggie Dog Lunch outside the Longworth House Office Building July 21, 2021 in Washington, DC. (Win McNamee/Getty Images) "The Bible has long been used to justify all forms of oppression, so we've used ChatGPT to make it clear that a loving God would never endorse exploitation of or cruelty to animals," says PETA President Ingrid Newkirk. "It took God only six days to create the entire world, but we realized it would take us years to rewrite the whole Bible, which is why we've started with just the first book." PETA OFFERS TO PAY FOR OSCAR MAYER WIENERMOBILE'S STOLEN CATALYTIC CONVERTER IF IT BECOMES VEGAN MOBILE In Genesis Chapter 22, Abraham travels to the land of Moriah, where instead of slaughtering a ram to demonstrate his faith, he "befriends a gentle lamb to show his reverence and respect for God's creation."

In many humanitarian scenarios, translation into severely low resource languages often does not require a universal translation engine, but a dedicated text-specific translation engine. For example, healthcare records, hygienic procedures, government communication, emergency procedures and religious texts are all limited texts. While generic translation engines for all languages do not exist, translation of multilingually known limited texts into new, endangered languages may be possible and reduce human translation effort. We attempt to leverage translation resources from many rich resource languages to efficiently produce best possible translation quality for a well known text, which is available in multiple languages, in a new, severely low resource language. We examine two approaches: 1. best selection of seed sentences to jump start translations in a new language in view of best generalization to the remainder of a larger targeted text(s), and 2. we adapt large general multilingual translation engines from many other languages to focus on a specific text in a new, unknown language. We find that adapting large pretrained multilingual models to the domain/text first and then to the severely low resource language works best. If we also select a best set of seed sentences, we can improve average chrF performance on new test languages from a baseline of 21.9 to 50.7, while reducing the number of seed sentences to only around 1,000 in the new, unknown language.

Harnessing logical reasoning ability is a comprehensive natural language understanding endeavor. With the release of Generative Pretrained Transformer 4 (GPT-4), highlighted as "advanced" at reasoning tasks, we are eager to learn the GPT-4 performance on various logical reasoning tasks. This report analyses multiple logical reasoning datasets, with popular benchmarks like LogiQA and ReClor, and newly-released datasets like AR-LSAT. We test the multi-choice reading comprehension and natural language inference tasks with benchmarks requiring logical reasoning. We further construct a logical reasoning out-of-distribution dataset to investigate the robustness of ChatGPT and GPT-4. We also make a performance comparison between ChatGPT and GPT-4. Experiment results show that ChatGPT performs significantly better than the RoBERTa fine-tuning method on most logical reasoning benchmarks. With early access to the GPT-4 API we are able to conduct intense experiments on the GPT-4 model. The results show GPT-4 yields even higher performance on most logical reasoning datasets. Among benchmarks, ChatGPT and GPT-4 do relatively well on well-known datasets like LogiQA and ReClor. However, the performance drops significantly when handling newly released and out-of-distribution datasets. Logical reasoning remains challenging for ChatGPT and GPT-4, especially on out-of-distribution and natural language inference datasets. We release the prompt-style logical reasoning datasets as a benchmark suite and name it LogiEval.

Target detection and recognition is a very challenging task in a wireless environment where a multitude of objects are located, whether to effectively determine their positions or to identify them and predict their moves. In this work, we propose a new method based on a convolutional neural network (CNN) to estimate the range and velocity of moving targets directly from the range-Doppler map of the detected signals. We compare the obtained results to the two dimensional (2D) periodogram, and to the similar state of the art methods, 2DResFreq and VGG-19 network and show that the estimation process performed with our model provides better estimation accuracy of range and velocity index in different signal to noise ratio (SNR) regimes along with a reduced prediction time. Afterwards, we assess the performance of our proposed algorithm using the peak signal to noise ratio (PSNR) which is a relevant metric to analyse the quality of an output image obtained from compression or noise reduction. Compared to the 2D-periodogram, 2DResFreq and VGG-19, we gain 33 dB, 21 dB and 10 dB, respectively, in terms of PSNR when SNR = 30 dB.

Climate change may be classified as the most important environmental problem that the Earth is currently facing, and affects all living species on Earth. Given that air-quality monitoring stations are typically ground-based their abilities to detect pollutant distributions are often restricted to wide areas. Satellites however have the potential for studying the atmosphere at large; the European Space Agency (ESA) Copernicus project satellite, "Sentinel-5P" is a newly launched satellite capable of measuring a variety of pollutant information with publicly available data outputs. This paper seeks to create a multi-modal machine learning model for predicting air-quality metrics where monitoring stations do not exist. The inputs of this model will include a fusion of ground measurements and satellite data with the goal of highlighting pollutant distribution and motivating change in societal and industrial behaviors. A new dataset of European pollution monitoring station measurements is created with features including $\textit{altitude, population, etc.}$ from the ESA Copernicus project. This dataset is used to train a multi-modal ML model, Air Quality Network (AQNet) capable of fusing these various types of data sources to output predictions of various pollutants. These predictions are then aggregated to create an "air-quality index" that could be used to compare air quality over different regions. Three pollutants, NO$_2$, O$_3$, and PM$_{10}$, are predicted successfully by AQNet and the network was found to be useful compared to a model only using satellite imagery. It was also found that the addition of supporting data improves predictions. When testing the developed AQNet on out-of-sample data of the UK and Ireland, we obtain satisfactory estimates though on average pollution metrics were roughly overestimated by around 20\%.

This study investigates the potential of an ambulatory device that incorporates Large Language Models (LLMs) in cadence with other specialized ML models to assess anemia severity in sickle cell patients in real time. The device would rely on sensor data that measures angiogenic material levels to assess anemia severity, providing real-time information to patients and clinicians to reduce the frequency of vaso-occlusive crises because of the early detection of anemia severity, allowing for timely interventions and potentially reducing the likelihood of serious complications. The main challenges in developing such a device are the creation of a reliable non-invasive tool for angiogenic level assessment, a biophysics model and the practical consideration of an LLM communicating with emergency personnel on behalf of an incapacitated patient. A possible system is proposed, and the limitations of this approach are discussed.

There is a broad consensus that news media outlets incorporate ideological biases in their news articles. However, prior studies on measuring the discrepancies among media outlets and further dissecting the origins of thematic differences suffer from small sample sizes and limited scope and granularity. In this study, we use a large dataset of 1.8 million news headlines from major U.S. media outlets spanning from 2014 to 2022 to thoroughly track and dissect the fine-grained thematic discrepancy in U.S. news media. We employ multiple correspondence analysis (MCA) to quantify the fine-grained thematic discrepancy related to four prominent topics - domestic politics, economic issues, social issues, and foreign affairs in order to derive a more holistic analysis. Additionally, we compare the most frequent $n$-grams in media headlines to provide further qualitative insights into our analysis. Our findings indicate that on domestic politics and social issues, the discrepancy can be attributed to a certain degree of media bias. Meanwhile, the discrepancy in reporting foreign affairs is largely attributed to the diversity in individual journalistic styles. Finally, U.S. media outlets show consistency and high similarity in their coverage of economic issues.

For monitoring crises, political events are extracted from the news. The large amount of unstructured full-text event descriptions makes a case-by-case analysis unmanageable, particularly for low-resource humanitarian aid organizations. This creates a demand to classify events into event types, a task referred to as event coding. Typically, domain experts craft an event type ontology, annotators label a large dataset and technical experts develop a supervised coding system. In this work, we propose PR-ENT, a new event coding approach that is more flexible and resource-efficient, while maintaining competitive accuracy: first, we extend an event description such as "Military injured two civilians'' by a template, e.g. "People were [Z]" and prompt a pre-trained (cloze) language model to fill the slot Z. Second, we select answer candidates Z* = {"injured'', "hurt"...} by treating the event description as premise and the filled templates as hypothesis in a textual entailment task. This allows domain experts to draft the codebook directly as labeled prompts and interpretable answer candidates. This human-in-the-loop process is guided by our interactive codebook design tool. We evaluate PR-ENT in several robustness checks: perturbing the event description and prompt template, restricting the vocabulary and removing contextual information.