The focus of the present paper is to investigate the following questions: to what extent syntactic features capture phylogenetic relationships and to what extent Markov models are a viable assumption for phylogenetic reconstruction based on syntactic features. For the second, we also consider an alternative that we argue approximates the infinite site evolutionary model. These questions are motivated by the fact that at both lexical and syntactic level, Markov processes are commonly assumed to underlie computational models of language change; for instance, within the Principles and Parameters setting relevant here, Niyogi and Berwick (1997) developed models of language acquisition and language change based on a Markov process in a space of syntactic parameters. In this paper we focus only on language change processes, viewed through the lens of phylogenetic trees of language families. While the model we consider are not directly related to models of language acquisition and parameter setting, the historical changes of syntax within and across language families, through the modification of syntactic parameters, can be seen as an effect of such underlying dynamics.

This paper presents Domain-Specific Sub-network (DoSS). It uses a set of masks obtained through pruning to define a sub-network for each domain and finetunes the sub-network parameters on domain data. This performs very closely and drastically reduces the number of parameters compared to finetuning the whole network on each domain. Also a method to make masks unique per domain is proposed and shown to greatly improve the generalization to unseen domains. In our experiments on German to English machine translation the proposed method outperforms the strong baseline of continue training on multi-domain (medical, tech and religion) data by 1.47 BLEU points. Also continue training DoSS on new domain (legal) outperforms the multi-domain (medical, tech, religion, legal) baseline by 1.52 BLEU points.

There is growing interest in concept-based models (CBMs) that combine high-performance and interpretability by acquiring and reasoning with a vocabulary of high-level concepts. A key requirement is that the concepts be interpretable. Existing CBMs tackle this desideratum using a variety of heuristics based on unclear notions of interpretability, and fail to acquire concepts with the intended semantics. We address this by providing a clear definition of interpretability in terms of alignment between the model's representation and an underlying data generation process, and introduce GlanceNets, a new CBM that exploits techniques from disentangled representation learning and open-set recognition to achieve alignment, thus improving the interpretability of the learned concepts. We show that GlanceNets, paired with concept-level supervision, achieve better alignment than state-of-the-art approaches while preventing spurious information from unintendedly leaking into the learned concepts.

This paper presents our submission to the 2022 edition of the CASE 2021 shared task 1, subtask 4. The EventGraph system adapts an end-to-end, graph-based semantic parser to the task of Protest Event Extraction and more specifically subtask 4 on event trigger and argument extraction. We experiment with various graphs, encoding the events as either "labeled-edge" or "node-centric" graphs. We show that the "node-centric" approach yields best results overall, performing well across the three languages of the task, namely English, Spanish, and Portuguese. EventGraph is ranked 3rd for English and Portuguese, and 4th for Spanish. Our code is available at: https://github.com/huiling-y/eventgraph_at_case

We present Burst2Vec, our multi-task learning approach to predict emotion, age, and origin (i.e., native country/language) from vocal bursts. Burst2Vec utilises pre-trained speech representations to capture acoustic information from raw waveforms and incorporates the concept of model debiasing via adversarial training. Our models achieve a relative 30 % performance gain over baselines using pre-extracted features and score the highest amongst all participants in the ICML ExVo 2022 Multi-Task Challenge.

Neural scaling laws define a predictable relationship between a model's parameter count and its performance after training in the form of a power law. However, most research to date has not explicitly investigated whether scaling laws can be used to accelerate model development. In this work, we perform such an empirical investigation across a wide range of language understanding tasks, starting from models with as few as 10K parameters, and evaluate downstream performance across 9 language understanding tasks. We find that scaling laws emerge at finetuning time in some NLP tasks, and that they can also be exploited for debugging convergence when training large models. Moreover, for tasks where scaling laws exist, they can be used to predict the performance of larger models, which enables effective model selection. However, revealing scaling laws requires careful hyperparameter tuning and multiple runs for the purpose of uncertainty estimation, which incurs additional overhead, partially offsetting the computational benefits.

Us humans have always worked towards making our lives easier and better, and this constant struggle to achieve something better worked as bliss for humans. Isn't it so fascinating to look back at our cave-devilling ancestors and realise how far we have advanced as humans? We went through various milestones to achieve the technology we have today. As we further surpassed in technology, we stumbled upon exploring artificial intelligence. The artificial intelligence (AI) we have today is in a golden age right now. Every industry is undergoing a sea change due to AI's inflection point. Specific applications of AI have already been discussed in great detail. Consider this post a complete guide on the practical use and foreseeing of artificial intelligence and how it impacts our lives. As a starter, I offer five bold predictions about how artificial intelligence will fundamentally alter our economy and society in the next decade.

Ai-Da, the world's first ultra-realistic humanoid AI robot artist, has made history once again due to her appearance in the House of Lords, the second chamber of the UK Parliament, where she addressed the question of whether creativity is under attack in today's ever-changing, technology-driven world. Ai-Da's address to members of the House of Lords Communications and Digital Committee was part of the House of Lords' inquiry into the future of the creative industries. During her speech, she explored the topic of AI and how this new technology is pushing the boundaries of how we think about creativity. We are entering a new era of machine creativity that presents new possibilities of creativity and technology beyond what humans can do. Ai-Da's creativity, which is driven by AI, sparks an in-depth conversation on what it means to be human in a post-human society, at a time when technology is fostering creativity like never before.

The use by Russia of Iranian drones in its war against Ukraine makes clear the weaknesses of its domestic industry and Tehran's growing claim on the market for unmanned aircraft, experts say. Washington believes Iran has delivered hundreds of drones, which Ukrainian officials say are now being used in strikes like those launched against cities and energy infrastructure on Monday. So far two models of Iranian drone have been identified in Ukraine's skies, built for two different purposes. One of them, the Shahed 136, is a relatively low-cost "kamikaze drone" that can be programmed to fly automatically to a set of GPS coordinates with a payload of explosives. "It flies quite low, striking a target that must be stationary at a range of a few hundred kilometres," said Pierre Grasser, a researcher tied to Paris' Sorbonne University.

Adrienne Williams and Milagros Miceli are researchers at the Distributed AI Research (DAIR) Institute. Timnit Gebru is the institute's founder and executive director. She was previously co-lead of the Ethical AI research team at Google. The public's understanding of artificial intelligence (AI) is largely shaped by pop culture -- by blockbuster movies like "The Terminator" and their doomsday scenarios of machines going rogue and destroying humanity. This kind of AI narrative is also what grabs the attention of news outlets: a Google engineer claiming that its chatbot was sentient was among the most discussed AI-related news in recent months, even reaching Stephen Colbert's millions of viewers.