I am very excited to announce that my Cryptics book is now available in U.S and many other countries on Amazon. This is a perfect holiday gift for students, life-long learners, professionals, entrepreneurs and anyone interested in getting introduced to amazing math and crypto concepts in a fun way. Cryptics is a hard science fiction novel and a technological thriller aimed at introducing some of the most amazing concepts of cryptography, blockchain, zero-knowledge proofs, artificial intelligence and mathematics in a fun, exciting and a memorable way. It is a story of five middle school students, who meet at a talk given by Dr. Shiva Kintali, a billionaire mathematician and a technologist. Doc brings their attention to a puzzle, whose solution is worth $10 Billion.

Model-based reinforcement learning (RL) is a sample-efficient way of learning complex behaviors by leveraging a learned single-step dynamics model to plan actions in imagination. However, planning every action for long-horizon tasks is not practical, akin to a human planning out every muscle movement. Instead, humans efficiently plan with high-level skills to solve complex tasks. From this intuition, we propose a Skill-based Model-based RL framework (SkiMo) that enables planning in the skill space using a skill dynamics model, which directly predicts the skill outcomes, rather than predicting all small details in the intermediate states, step by step. For accurate and efficient long-term planning, we jointly learn the skill dynamics model and a skill repertoire from prior experience. We then harness the learned skill dynamics model to accurately simulate and plan over long horizons in the skill space, which enables efficient downstream learning of long-horizon, sparse reward tasks. Experimental results in navigation and manipulation domains show that SkiMo extends the temporal horizon of model-based approaches and improves the sample efficiency for both model-based RL and skill-based RL. Code and videos are available at https://clvrai.com/skimo

This document describes strategies for using Artificial Intelligence (AI) to predict some journal article scores in future research assessment exercises. Five strategies have been assessed. These are summarised here for completeness, but we recommend that AI predictions are not used to help make scoring decisions yet but are further explored through pilot testing in the next REF or REF replacement. The pilot testing should assess whether using AI predictions and prediction probabilities alongside, or instead of, bibliometric data would be helpful for any UoAs. For example, depending on UoA, AI predictions may be used to help mop up difficult scoring decisions near the end of the assessment period, to gain interdisciplinary input, as a tiebreaker in the way that bibliometrics are currently sometimes used, or to cross check the final scores.

The success of graph neural networks on graph-based web mining highly relies on abundant human-annotated data, which is laborious to obtain in practice. When only few labeled nodes are available, how to improve their robustness is a key to achieve replicable and sustainable graph semi-supervised learning. Though self-training has been shown to be powerful for semi-supervised learning, its application on graph-structured data may fail because (1) larger receptive fields are not leveraged to capture long-range node interactions, which exacerbates the difficulty of propagating feature-label patterns from labeled nodes to unlabeled nodes; and (2) limited labeled data makes it challenging to learn well-separated decision boundaries for different node classes without explicitly capturing the underlying semantic structure. To address the challenges of capturing informative structural and semantic knowledge, we propose a new graph data augmentation framework, AGST (Augmented Graph Self-Training), which is built with two new (i.e., structural and semantic) augmentation modules on top of a decoupled GST backbone. In this work, we investigate whether this novel framework can learn a robust graph predictive model under the low-data context. We conduct comprehensive evaluations on semi-supervised node classification under different scenarios of limited labeled-node data. The experimental results demonstrate the unique contributions of the novel data augmentation framework for node classification with few labeled data.

The goal of building dialogue agents that can converse with humans naturally has been a long-standing dream of researchers since the early days of artificial intelligence. The well-known Turing Test proposed to judge the ultimate validity of an artificial intelligence agent on the indistinguishability of its dialogues from humans'. It should come as no surprise that human-level dialogue systems are very challenging to build. But, while early effort on rule-based systems found limited success, the emergence of deep learning enabled great advance on this topic. In this thesis, we focus on methods that address the numerous issues that have been imposing the gap between artificial conversational agents and human-level interlocutors. These methods were proposed and experimented with in ways that were inspired by general state-of-the-art AI methodologies. But they also targeted the characteristics that dialogue systems possess.

Reasoning is central to human intelligence. However, fallacious arguments are common, and some exacerbate problems such as spreading misinformation about climate change. In this paper, we propose the task of logical fallacy detection, and provide a new dataset (Logic) of logical fallacies generally found in text, together with an additional challenge set for detecting logical fallacies in climate change claims (LogicClimate). Detecting logical fallacies is a hard problem as the model must understand the underlying logical structure of the argument. We find that existing pretrained large language models perform poorly on this task. In contrast, we show that a simple structure-aware classifier outperforms the best language model by 5.46% on Logic and 4.51% on LogicClimate. We encourage future work to explore this task as (a) it can serve as a new reasoning challenge for language models, and (b) it can have potential applications in tackling the spread of misinformation. Our dataset and code are available at https://github.com/causalNLP/logical-fallacy

This literature review identifies indicators that associate with higher impact or higher quality research from article text (e.g., titles, abstracts, lengths, cited references and readability) or metadata (e.g., the number of authors, international or domestic collaborations, journal impact factors and authors' h-index). This includes studies that used machine learning techniques to predict citation counts or quality scores for journal articles or conference papers. The literature review also includes evidence about the strength of association between bibliometric indicators and quality score rankings from previous UK Research Assessment Exercises (RAEs) and REFs in different subjects and years and similar evidence from other countries (e.g., Australia and Italy). In support of this, the document also surveys studies that used public datasets of citations, social media indictors or open review texts (e.g., Dimensions, OpenCitations, Altmetric.com and Publons) to help predict the scholarly impact of articles. The results of this part of the literature review were used to inform the experiments using machine learning to predict REF journal article quality scores, as reported in the AI experiments report for this project. The literature review also covers technology to automate editorial processes, to provide quality control for papers and reviewers' suggestions, to match reviewers with articles, and to automatically categorise journal articles into fields. Bias and transparency in technology assisted assessment are also discussed.

Those who, like this columnist, spend too much time online will have noticed a kind of feeding frenzy over the past two weeks. The cause has been the release of an interesting chatbot – a software application capable of conducting an online conversation. The particular bot creating the fuss is ChatGPT, a prototype artificial intelligence (AI) chatbot that focuses on usability and dialogue and was developed by OpenAI, an AI research laboratory based in San Francisco. ChatGPT uses a large language model built via machine-learning methods and is based on OpenAI's GPT-3 model, which is capable of producing human-like text when given a prompt in natural language. It's an example of what has come to be called "generative AI": software that uses machine-learning algorithms to enable machines to generate artificial content – text, images, audio and video content based on its training data – in a way that might persuade a human user into believing that its outputs are "real".

Not only does ChatGPT write clear, compelling essays, but it can also conjure up its own personal ... [ ] details and embellishments that could up a students' chance of acceptance and would be difficult to verify. Forbes' full conversation with ChatGPT, OpenAI's newest natural language model, is pasted below. Each of the college admissions essays took less than 10 minutes to complete. Read our story about ChatGPT's capacity to write college applications here. Forbes: Hi GPT, I'd like you to write a college application essay as if you were an 18-year-old high school senior whose parents are from Bangalore, India but who now own a restaurant in Newton, Mass. He is a competitive swimmer, and in 10th grade he broke his shoulder. He is interested in majoring in business.

Technology is changing the way we learn, work, and live. Many of the current jobs will become obsolete in another 5 -10 years with automation and 50% of the new jobs will be those which don't exist today. Understanding the language of the machines will be a standard that is as important as having English literacy or even a native-tongue literacy. And, the future offers a humongous number of opportunities for the people who know how to code. Cyber Square is a unique platform for students and teachers to get training on the latest technologies like Artificial Intelligence, Robotics, IoT, etc. Cyber Square curriculum and platform will help kids to develop their own projects using the latest technologies through schools, we are providing coding for schools.