Partly automated creation of interlinear glossed text (IGT) has the potential to assist in linguistic documentation. We argue that LLMs can make this process more accessible to linguists because of their capacity to follow natural-language instructions. We investigate the effectiveness of a retrieval-based LLM prompting approach to glossing, applied to the seven languages from the SIGMORPHON 2023 shared task. Our system beats the BERT-based shared task baseline for every language in the morpheme-level score category, and we show that a simple 3-best oracle has higher word-level scores than the challenge winner (a tuned sequence model) in five languages. In a case study on Tsez, we ask the LLM to automatically create and follow linguistic instructions, reducing errors on a confusing grammatical feature. Our results thus demonstrate the potential contributions which LLMs can make in interactive systems for glossing, both in making suggestions to human annotators and following directions.

We explore off-policy evaluation and learning (OPE/L) in contextual combinatorial bandits (CCB), where a policy selects a subset in the action space. For example, it might choose a set of furniture pieces (a bed and a drawer) from available items (bed, drawer, chair, etc.) for interior design sales. This setting is widespread in fields such as recommender systems and healthcare, yet OPE/L of CCB remains unexplored in the relevant literature. Typical OPE/L methods such as regression and importance sampling can be applied to the CCB problem, however, they face significant challenges due to high bias or variance, exacerbated by the exponential growth in the number of available subsets. To address these challenges, we introduce a concept of factored action space, which allows us to decompose each subset into binary indicators. This formulation allows us to distinguish between the ''main effect'' derived from the main actions, and the ''residual effect'', originating from the supplemental actions, facilitating more effective OPE. Specifically, our estimator, called OPCB, leverages an importance sampling-based approach to unbiasedly estimate the main effect, while employing regression-based approach to deal with the residual effect with low variance. OPCB achieves substantial variance reduction compared to conventional importance sampling methods and bias reduction relative to regression methods under certain conditions, as illustrated in our theoretical analysis. Experiments demonstrate OPCB's superior performance over typical methods in both OPE and OPL.

In this paper, we carefully investigate how we can use multiple different Natural Language Processing techniques and methods in order to automatically recognize the main actions in sports events. We aim to extract insights by analyzing live sport commentaries from different sources and by classifying these major actions into different categories. We also study if sentiment analysis could help detect these main actions.

Last night as I was trying to sleep, I was thinking about robots and how they would operate in the future. I had wondered how robots would work with us but more importantly, how robots would work with each other because many robots may be built by different companies and may not be on the same communication line. After sleeping on it, I did some research the next day and came across a new article written by UIUC that spoke almost exactly to my curiosity. A few researchers at the Grainger College of Engineering developed a sort of artificial intelligence that used "multi-agent reinforcement learning", a term that I have not heard of before. After diving deeper into it, the machine basically learns from its past experiences and builds a training method that rewards and punishes certain decisions made by the machine.

After a while, we just created a layout of the scene and took it from there." It's a simple story to follow, and most of the action takes place right in front of you. The viewer is free, however, to look elsewhere, and if you're using the Spotlight Stories app, the movie will adapt accordingly. If you look at the musicians, for instance, the main action -- Méliès and his wife -- will wait off-screen until you turn your head back. That's possible because of some special Google software called the Spotlight Stories Editor, which allows for nodal-based logic, similar to what's used in a video game. "It's immersive theater much more than it is a film." Mark Davies, a CG supervisor at Nexus Studios, says it works like punch-drunk-style theater. "It's immersive theater much more than it is a film," he said. "Because the actor is there, and if they see that you're staring at the floor, the ceiling or something else, they'll wait for you to turn around, and go, 'Oh, yes, I can begin acting again.'"