Meta-learning is a promising strategy for learning to efficiently learn using data gathered from a distribution of tasks. However, the meta-learning literature thus far has focused on the task segmented setting, where at train-time, offline data is assumed to be split according to the underlying task, and at test-time, the algorithms are optimized to learn in a single task. In this work, we enable the application of generic meta-learning algorithms to settings where this task segmentation is unavailable, such as continual online learning with unsegmented time series data.

We propose a heterogeneous meta-learning method that trains a model on tasks with various attribute spaces, such that it can solve unseen tasks whose attribute spaces are different from the training tasks given a few labeled instances. Although many meta-learning methods have been proposed, they assume that all training and target tasks share the same attribute space, and they are inapplicable when attribute sizes are different across tasks. Our model infers latent representations of each attribute and each response from a few labeled instances using an inference network. Then, responses of unlabeled instances are predicted with the inferred representations using a prediction network. The attribute and response representations enable us to make predictions based on the task-specific properties of attributes and responses even when attribute and response sizes are different across tasks. In our experiments with synthetic datasets and 59 datasets in OpenML, we demonstrate that our proposed method can predict the responses given a few labeled instances in new tasks after being trained with tasks with heterogeneous attribute spaces.

Large Language Models (LLMs) trained on massive corpora have shown remarkable success in knowledge-intensive tasks. Y et, most of them rely on pre-stored knowledge. Inducing new general knowledge from a specific environment and performing reasoning with the acquired knowledge-- situated inductive reasoning, is crucial and challenging for machine intelligence. In this paper, we design Mars, an interactive environment devised for situated inductive reasoning. It introduces counter-commonsense game mechanisms by modifying terrain, survival setting and task dependency while adhering to certain principles.

Abstract: This paper presents an in-depth analysis of the performance of seven different Large Language Models (LLMs) in solving a diverse set of math advanced calculus problems. The study aims to evaluate these models' accuracy, reliability, and problem-solving capabilities, including ChatGPT 4o, Gemini Advanced with 1.5 Pro, Copilot Pro, Claude 3.5 Sonnet, Meta AI, Mistral AI, and Perplexity. The assessment was conducted through a series of thirty-two test problems, encompassing a total of 320 points. The problems covered a wide range of topics, from vector calculations and geometric interpretations to integral evaluations and optimization tasks. The results highlight significant trends and patterns in the models' performance, revealing both their strengths and weaknesses - for instance, models like ChatGPT 4o and Mistral AI demonstrated consistent accuracy across various problem types, indicating their robustness and reliability in mathematical problem-solving, while models such as Gemini Advanced with 1.5 Pro and Meta AI exhibited specific weaknesses, particularly in complex problems involving integrals and optimization, suggesting areas for targeted improvements. The study also underscores the importance of re-prompting in achieving accurate solutions, as seen in several instances where models initially provided incorrect answers but corrected them upon re-prompting. Overall, this research provides valuable insights into the current capabilities and limitations of LLMs in the domain of math calculus, with the detailed analysis of each model's performance on specific problems offering a comprehensive understanding of their strengths and areas for improvement, contributing to the ongoing development and refinement of LLM technology. The findings are particularly relevant for educators, researchers, and developers seeking to leverage LLMs for educational and practical applications in mathematics.

Enterprises are quickly shifting their IT help desk strategies away from one where every employee's issue or request requires human intervention to one that leverages artificial intelligence (AI)/natural language processing (NLP) for automated resolution. One area that the enterprise service management (ESM) market is now focusing on is the automation of tasks (e.g., fulfill a service request, create a new mailing list, schedule PTO, reserve guest desk). Once this problem was tackled, customers looked to automate virtually anything an employee could ask for -- essentially becoming a system of engagement for issues and requests. One of the most complex parts of creating automation with virtual support agents is to connect the automation to the human language. As the trend toward intelligent automation is moving well beyond IT to include HR and more, it's important that virtual support agent platforms enable organizations to accomplish tasks such as creating simple or complex integrations with virtually any REST API-enabled system.

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - August 3. Join AI and data leaders for insightful talks and exciting networking opportunities. In the post-pandemic future of work-from-anywhere, will our desktops, tablets and phones become de facto reporters of how well we perform our jobs? We're not anywhere close to this yet, but the spread-out nature of how work is getting done here in 2022 may be begging for closer and more automated supervision of the workforce, which some managers undoubtedly will want to investigate. That intelligent supervision will certainly include extensive use of AI and ML. One of the first tools to use some of this functionality may well be San Diego, California-based ClickUp, which makes a customizable, SaaS-based workplace productivity platform that serves all departments across an organization.

Are you looking to land a top-paying job in Data Science? Or are you a seasoned AI practitioner who want to take your career to the next level? Or are you an aspiring entrepreneur who wants to maximize business revenue with Data Science and Artificial Intelligence? If the answer is yes to any of these questions, then this course is for you! Data Science is one of the hottest tech fields to be in right now!

Can world knowledge learned by large language models (LLMs) be used to act in interactive environments? In this paper, we investigate the possibility of grounding high-level tasks, expressed in natural language (e.g. "make breakfast"), to a chosen set of actionable steps (e.g. "open fridge"). While prior work focused on learning from explicit step-by-step examples of how to act, we surprisingly find that if pre-trained LMs are large enough and prompted appropriately, they can effectively decompose high-level tasks into low-level plans without any further training. However, the plans produced naively by LLMs often cannot map precisely to admissible actions. We propose a procedure that conditions on existing demonstrations and semantically translates the plans to admissible actions. Our evaluation in the recent VirtualHome environment shows that the resulting method substantially improves executability over the LLM baseline. The conducted human evaluation reveals a trade-off between executability and correctness but shows a promising sign towards extracting actionable knowledge from language models. Website at https://huangwl18.github.io/language-planner

New developments in automation, hardware, model development, and more that will shape AI in 2020. Roger Magoulas, VP of Radar at O'Reilly takes a look at the new developments in automation, hardware, tools, model development, and more that will shape (or accelerate) AI in 2020. We see the AI space poised for an acceleration in adoption, driven by more sophisticated AI models being put in production, specialised hardware that increases AI's capacity to provide quicker results based on larger datasets, simplified tools that democratise access to the entire AI stack, small tools that enables AI on nearly any device, and cloud access to AI tools that allow access to AI resources from anywhere. Integrating data from many sources, complex business and logic challenges, and competitive incentives to make data more useful all combine to elevate AI and automation technologies from optional to required. And AI processes have unique capabilities that can address an increasingly diverse array of automation tasks--tasks that defy what traditional procedural logic and programming can handle, for example, image recognition, summarisation, labeling, complex monitoring, and response.

Delane Parnell is the cofounder and CEO of PlayVS. If there's ever a constant in the flourishing world of esports, it's that enthusiasm often outpaces the necessary infrastructure to match it. In particular, high school students and teachers who hope to participate in competitive gaming must self-organize without the structure of an official body. Delane Parnell's high school science teacher was someone who took it upon themselves to organize a gaming club for students. He provided the equipment, he kept track of stats and even awarded trophies for the myriad of games they played.