Contextual predictability shapes both the form and choice of words in online language production. The effects of the predictability of a word given its previous context are generally well-understood in both production and comprehension, but studies of naturalistic production have also revealed a poorly-understood backward predictability effect of a word given its future context, which may be related to future planning. Here, in two studies of naturalistic speech corpora, we investigate backward predictability effects using improved measures and more powerful language models, introducing a new principled and conceptually motivated information-theoretic predictability measure that integrates predictability from both the future and the past context. Our first study revisits classic predictability effects on word duration. Our second study investigates substitution errors within a generative framework that independently models the effects of lexical, contextual, and communicative factors on word choice, while predicting the actual words that surface as speech errors. We find that our proposed conceptually-motivated alternative to backward predictability yields qualitatively similar effects across both studies. Through a fine-grained analysis of substitution errors, we further show that different kinds of errors are suggestive of how speakers prioritize form, meaning, and context-based information during lexical planning. Together, these findings illuminate the functional roles of past and future context in how speakers encode and choose words, offering a bridge between contextual predictability effects and the mechanisms of sentence planning.

Imagine AI assistants that enhance conversations without interrupting them: quietly providing relevant information during a medical consultation, seamlessly preparing materials as teachers discuss lesson plans, or unobtrusively scheduling meetings as colleagues debate calendars. While modern conversational LLM agents directly assist human users with tasks through a chat interface, we study this alternative paradigm for interacting with LLM agents, which we call "overhearing agents". Rather than demanding the user's attention, overhearing agents continuously monitor ambient activity and intervene only when they can provide contextual assistance. In this paper, we present the first analysis of overhearing LLM agents as a distinct paradigm in human-AI interaction and establish a taxonomy of overhearing agent interactions and tasks grounded in a survey of works on prior LLM-powered agents and exploratory HCI studies. Based on this taxonomy, we create a list of best practices for researchers and developers building overhearing agent systems. Finally, we outline the remaining research gaps and reveal opportunities for future research in the overhearing paradigm.

The growing number of pretrained models in Machine Learning (ML) presents significant challenges for practitioners. Given a new dataset, they need to determine the most suitable deep learning (DL) pipeline, consisting of the pretrained model and the hyperparameters for finetuning to it. Moreover, as models grow in scale, the increasing reliance on real-world data poses a bottleneck for training and requires leveraging data more effectively. Addressing the first challenge often involves manual model selection and hyperparameter tuning. At the same time, as models grow larger and more and more of the available human-generated data is being used for training, data augmentation and synthetic data become critical elements. Automated machine learning offers a path to address these challenges but is traditionally designed for tabular data and classical ML methods. This dissertation adopts meta-learning to extend automated machine learning to the deep learning domain. We propose empirical approaches to automate DL pipeline selection for Computer Vision tasks using prior task knowledge to learn surrogate models for pipeline ranking. Extending these methods to the language domain, we learn to finetune large language models. As a result, we show that our approach can outperform finetuning foundation models. Additionally, we meta-learn data augmentation and synthetic data to enhance performance in up-stream and down-stream tasks. We empirically show the underestimated importance of data augmentation when using Self-Supervised Learning and meta-learn advanced data augmentation strategies. Leveraging synthetic data, we also propose to meta-learn neural synthetic data generators as proxies for Reinforcement Learning (RL) environments. Additionally, we learn a multiple-environment world model in an in-context learning fashion by purely using synthetic, randomly sampled data.

Psycholinguistic research suggests that humans may build a representation of linguistic input that is 'good-enough' for the task at hand. This study examines what architectural features make language models learn human-like good-enough language processing. We focus on the number of layers and self-attention heads in Transformers. We create a good-enough language processing (GELP) evaluation dataset (7,680 examples), which is designed to test the effects of two plausibility types, eight construction types, and three degrees of memory cost on language processing. To annotate GELP, we first conduct a crowdsourcing experiment whose design follows prior psycholinguistic studies. Our model evaluation against the annotated GELP then reveals that the full model as well as models with fewer layers and/or self-attention heads exhibit a good-enough performance. This result suggests that models with shallower depth and fewer heads can learn good-enough language processing.

Making perfume is an art that can be traced back to ancient Greece but now modern-day perfumiers are beginning to look beyond their noses to develop the scents most likely to appeal to us. They are, instead, turning to AI. Perfumes can now be designed to trigger emotional responses using ingredients known as neuroscents – odours shown by biometric measures to arouse different positive feelings such as calm, euphoria or sleepiness. Hugo Ferreira, a researcher at the Institute of Biophysics and Biomedical Engineering in Lisbon, is mapping brain activity and response to perfumes to build a database of neuroscents. He says the sense of smell is fascinating. "With sight and hearing, you can imagine the face of a loved one or favourite tune. It's hard to imagine a smell even though [it] can provoke a torrent of emotions and memories."

New York City grocers are expressing outrage over a push by city council members to ban facial recognition technology stores rely on to deter shoplifting due to concerns of racial discrimination. Ferreira Foodtown CEO Jason Ferreira joined "Fox & Friends" Tuesday to call out the suggestion as thefts continue to rock businesses in the Big Apple. Ferreira, who has been in business for over 45 years, said the shoplifting has never been worse. "It's not only people that are doing it professionally. We have people that are doing it just because they can get away with it. And the gamut runs from children to people that are older."

This paper presents a new approach for controlling emotion in symbolic music generation with Monte Carlo Tree Search. We use Monte Carlo Tree Search as a decoding mechanism to steer the probability distribution learned by a language model towards a given emotion. At every step of the decoding process, we use Predictor Upper Confidence for Trees (PUCT) to search for sequences that maximize the average values of emotion and quality as given by an emotion classifier and a discriminator, respectively. We use a language model as PUCT's policy and a combination of the emotion classifier and the discriminator as its value function. To decode the next token in a piece of music, we sample from the distribution of node visits created during the search. We evaluate the quality of the generated samples with respect to human-composed pieces using a set of objective metrics computed directly from the generated samples. We also perform a user study to evaluate how human subjects perceive the generated samples' quality and emotion. We compare PUCT against Stochastic Bi-Objective Beam Search (SBBS) and Conditional Sampling (CS). Results suggest that PUCT outperforms SBBS and CS in almost all metrics of music quality and emotion.

This project aims to compose background music in real-time for tabletop role-playing games. To accomplish this goal, we propose a system called MTG that listens to players' speeches in order to recognize the context of the current scene and generate background music to match the scene. A speech recognition system is used to transcribe players' speeches to text and a supervised learning algorithm detects when scene transitions take place. In its current version, a scene transition occurs whenever the emotional state of the narrative changes. Moreover, the background music is not generated, but selected based on its emotion from a library of hand-authored pieces. As future work, we plan to generate the background music considering the current scene context and the probability of scene transition. We also consider to retrieve more information from the narrative to detect scene transitions, such as the scene's location and time of the day as well as actions taken by characters.

LAS VEGAS – At 18 months old, Chris Rodriguez was diagnosed with heart failure and required a transplant. "My heart was pumping too much blood and my organs couldn't deal with it," said Rodriguez, now 14. "Most of my life, I've just been in and out of the hospital. More than 100,000 people are on the national transplant waiting list, and approximately 17 people die each day waiting to receive an organ transplant, according to the Health Resources and Services Administration. COVID-19 has complicated the situation even more, as travel restrictions and fewer commercial flights have made it difficult to transplant organs and highlighted the need for alternative travel methods to deliver vital organs. Nevada Donor Network partnered with MissionGo to test drones to deliver vital organs. "As a result of the COVID-19 pandemic, we have been subjected to fewer commercial flights to be able to transport organs for transplantation," Joe Ferreira, Nevada Donor Network CEO said. "We've had to look ...

Anyone can tell an eagle from an ostrich. It takes a skilled birdwatcher to tell a chipping sparrow from a house sparrow from an American tree sparrow. Now researchers are using AI to take this to the next level -- identifying individual birds. André Ferreira, a Ph.D. student at France's Centre for Functional and Evolutionary Ecology, harnessed an NVIDIA GeForce RTX 2070 to train a powerful AI that identifies individual birds within the same species. It's the latest example of how deep learning has become a powerful tool for wildlife biologists studying a wide range of animals.