What does it mean for a generative AI model to be explainable? The emergent discipline of explainable AI (XAI) has made great strides in helping people understand discriminative models. Less attention has been paid to generative models that produce artifacts, rather than decisions, as output. Meanwhile, generative AI (GenAI) technologies are maturing and being applied to application domains such as software engineering. Using scenario-based design and question-driven XAI design approaches, we explore users' explainability needs for GenAI in three software engineering use cases: natural language to code, code translation, and code auto-completion. We conducted 9 workshops with 43 software engineers in which real examples from state-of-the-art generative AI models were used to elicit users' explainability needs. Drawing from prior work, we also propose 4 types of XAI features for GenAI for code and gathered additional design ideas from participants. Our work explores explainability needs for GenAI for code and demonstrates how human-centered approaches can drive the technical development of XAI in novel domains.

Ushio et al., 2021; Paranjape et al., 2021), multiword expression (MWE) identification (Espinosa-In this paper, we analyze zero-shot taxonomy In this paper, we evaluate LMs on TL benchmarks Taxonomy learning (TL) is the task of arranging using prompt-based and sentence-scoring domain terminologies into hierarchical structures techniques, and find not only that they are competitive where terms are nodes and edges denote is-a (hypernymic) with common approaches proposed in the relationships (Hwang et al., 2012). Domainspecific literature (which are typically supervised and/or concept generalization is at the core of human reliant on external resources), but that they achieve cognition (Yu et al., 2015), and a key enabler SoTa results in certain domains.

Kellogg's is perhaps best known for its breakfast cereals from its Corn Flakes to Frosties to Fruit Loops and more, and before the COVID-19 pandemic hit the APAC region, its dominant positioning was within the quick prework/school breakfast consumption occasion. But with the onset of the pandemic and most consumers having to stay home for both work and school, this consumption occasion became less attractive, and the company rightly realised a need to interpret consumer behaviour in a new light to pivot and operate accordingly. "We saw a big transformation where most consumers started to increase their at-home consumption significantly, which also included a lot more cooking at home as many found their inner chef – so we wanted to find a way to offer them more options to use our products in a way closer to their culture and habits," Kellogg's South East Asia Chief Marketing Officer Sanjib Bose told FoodNavigator-Asia. "Before COVID-19, we would have done this via normal interviews with consumers but realised this was now not possible, and we also wanted to go deeper to understand them better by finding out what they wanted directly from conversations they were leading online. "We felt that AI technology was a good way forward for this as it also helped us solve challenges such as language barriers, as most consumers will post on social media in different languages especially in Asia where there are so many different languages; and also to make sense of the huge volume of data from all of these online conversations." Kellogg's partnered with AI-specialist firm Ai Palette to do this, and apart from the use of AI to process data, the firm's technology is also language agnostic, which allowed it to help with the language challenge. "Our language agnostic model allowed us to gather and process data from various locations and across various diverse languages such as Bahasa Indonesia and Bahasa Melayu, whereas image analytics also helped to accurately identify data relevant to Kellogg's," Ai Palette CEO Somsubhra GanChoudhuri told us. "The big data found that now new consumption occasions for cereals have gone beyond breakfast – these are being used in proper recipes for cooking and baking, as a result of increased interest in home cooking and home baking during the pandemic.

Grindr is tightening privacy controls for the Olympic Village in Beijing. Bloomberg has learned the gay dating app is blocking people outside the Village from using the location-based Explore feature to find athletes in or near the area. The move is meant to protect athletes from harassment or persecution so they can "feel confident" connecting with each other during the Winter Olympic Games, Grindr for Equality director Jack Harrison-Quintana said. Anyone who uses Grindr inside the Village will see a pop-up telling them people outside the area can't browse the locale using Explore. "Your privacy is important to us," Grindr says in the alert.

The EU strikes again with a new set of regulations that take aim at the use of artificial intelligence (AI) to address the variety of risks associated with the societal adoption of AI. Like its sibling the General Data Protection Regulation (GDPR), the Artificial Intelligence Act (AIA) actually has teeth, with fines rising to €30 million, or 6% of global revenue. Is the answer to delete all your AI systems to minimize your risk to zero, or continue using AI for a competitive edge? Can you manage the recurring costs required to maintain compliance with the AIA even as the technology itself increases your bottomline? Take the famous UK pub chain JD Wetherspoon, founded by British businessman Tim Martin in 1979 who has been an outspoken critic of the EU and a Brexit campaigner. Their response to personal identifiable information (PII) protection, legislated by the GDPR in 2017, was to delete their entire CRM database.

One of the stars of the working-out-from-home boom is struggling. Peloton won't go quietly though and is making some big changes. The company will replace the CEO and co-founder, John Foley, who will become executive chairman, with former Spotify COO Barry McCarthy reportedly set to step into his shoes. While Foley is sticking around, the company is cutting around 2,800 corporate positions -- these won't include Peloton's instructors who lead its live classes. The company said in a press release about the lay-offs that its "monthly membership will be complimentary for impacted team members for an additional 12 months."

Dialogue understanding has always been a bottleneck for many conversational tasks, such as dialogue response generation and conversational question answering. To expedite the progress in this area, we introduce the task of conversational aspect sentiment analysis (CASA) that can provide useful fine-grained sentiment information for dialogue understanding and planning. Overall, this task extends the standard aspect-based sentiment analysis to the conversational scenario with several major adaptations. To aid the training and evaluation of data-driven methods, we annotate 3,000 chit-chat dialogues (27,198 sentences) with fine-grained sentiment information, including all sentiment expressions, their polarities and the corresponding target mentions. We also annotate an out-of-domain test set of 200 dialogues for robustness evaluation. Besides, we develop multiple baselines based on either pretrained BERT or self-attention for preliminary study. Experimental results show that our BERT-based model has strong performances for both in-domain and out-of-domain datasets, and thorough analysis indicates several potential directions for further improvements.

Pixel-level analysis of blood images plays a pivotal role in diagnosing blood-related diseases, especially Anaemia. These analyses mainly rely on an accurate diagnosis of morphological deformities like shape, size, and precise pixel counting. In traditional segmentation approaches, instance or object-based approaches have been adopted that are not feasible for pixel-level analysis. The convolutional neural network (CNN) model required a large dataset with detailed pixel-level information for the semantic segmentation of red blood cells in the deep learning domain. In current research work, we address these problems by proposing a multi-level deep convolutional encoder-decoder network along with two state-of-the-art healthy and Anaemic-RBC datasets. The proposed multi-level CNN model preserved pixel-level semantic information extracted in one layer and then passed to the next layer to choose relevant features. This phenomenon helps to precise pixel-level counting of healthy and anaemic-RBC elements along with morphological analysis. For experimental purposes, we proposed two state-of-the-art RBC datasets, i.e., Healthy-RBCs and Anaemic-RBCs dataset. Each dataset contains 1000 images, ground truth masks, relevant, complete blood count (CBC), and morphology reports for performance evaluation. The proposed model results were evaluated using crossmatch analysis with ground truth mask by finding IoU, individual training, validation, testing accuracies, and global accuracies using a 05-fold training procedure. This model got training, validation, and testing accuracies as 0.9856, 0.9760, and 0.9720 on the Healthy-RBC dataset and 0.9736, 0.9696, and 0.9591 on an Anaemic-RBC dataset. The IoU and BFScore of the proposed model were 0.9311, 0.9138, and 0.9032, 0.8978 on healthy and anaemic datasets, respectively.

Bayesian Neural Networks (BNNs) offer a mathematically grounded framework to quantify the uncertainty of model predictions but come with a prohibitive computation cost for both training and inference. In this work, we show a novel network architecture search (NAS) that optimizes BNNs for both accuracy and uncertainty while having a reduced inference latency. Different from canonical NAS that optimizes solely for in-distribution likelihood, the proposed scheme searches for the uncertainty performance using both in- and out-of-distribution data. Our method is able to search for the correct placement of Bayesian layer(s) in a network. In our experiments, the searched models show comparable uncertainty quantification ability and accuracy compared to the state-of-the-art (deep ensemble). In addition, the searched models use only a fraction of the runtime compared to many popular BNN baselines, reducing the inference runtime cost by $2.98 \times$ and $2.92 \times$ respectively on the CIFAR10 dataset when compared to MCDropout and deep ensemble.

This paper describes our submission to ICASSP 2022 Multi-channel Multi-party Meeting Transcription (M2MeT) Challenge. For Track 1, we propose several approaches to empower the clustering-based speaker diarization system to handle overlapped speech. Front-end dereverberation and the direction-of-arrival (DOA) estimation are used to improve the accuracy of speaker diarization. Multi-channel combination and overlap detection are applied to reduce the missed speaker error. A modified DOVER-Lap is also proposed to fuse the results of different systems. We achieve the final DER of 5.79% on the Eval set and 7.23% on the Test set. For Track 2, we develop our system using the Conformer model in a joint CTC-attention architecture. Serialized output training is adopted to multi-speaker overlapped speech recognition. We propose a neural front-end module to model multi-channel audio and train the model end-to-end. Various data augmentation methods are utilized to mitigate over-fitting in the multi-channel multi-speaker E2E system. Transformer language model fusion is developed to achieve better performance. The final CER is 19.2% on the Eval set and 20.8% on the Test set.