Neural circuits undergo developmental processes which can be influenced by experience. Here we explore a bio-inspired development process to form the connections in a network used for locality sensitive hashing.

AI has the potential to empower everyone to become more independent and self-sufficient. The increasing use of artificial intelligence (AI)-based technologies in everyday settings creates new opportunities to understand how disabled people might use these technologies [Glazko, 2023]. It also enables the development of new types of assistive technologies as well as new ways for people with disabilities to interact with technology in ways that are both simpler (for those who need things simpler) and more efficient and effective for those who cannot use the traditional interfaces effectively. AI has been rapidly taken up in almost all accessibility communities [Adnin 2024, Alharbi 2024, Jiang 2024, Bennett 2024, Valencia 2023]. Since becoming widely available to the public, Generative Artificial Intelligence (GAI) has steadily gained recognition for its potential as a valuable tool in the private sector and by government, as well as a tool for accessibility. Studies of blind and visually impaired individuals have found that they use GAI to'offload' cognitively demanding tasks and obtain personal help such as fashion advice (e.g., [Xie 2024]), and to create content or retrieve information [Adnin 2024]. A study of GAI use by neurodiverse users found GAI can both support and complicate tasks like code-switching, emotional regulation, and accessing information [Glazko, 2025]. A study of people who use AAC found it helpful for text input [Valencia 2023]. However there are concerns with a technology that is often based on probability and thus tends toward the most common case rather than those at the margins.

Implementing new features in repository-level codebases is a crucial application of code generation models. However, current benchmarks lack a dedicated evaluation framework for this capability. To fill this gap, we introduce FEA-Bench, a benchmark designed to assess the ability of large language models (LLMs) to perform incremental development within code repositories. We collect pull requests from 83 GitHub repositories and use rule-based and intent-based filtering to construct task instances focused on new feature development. Each task instance containing code changes is paired with relevant unit test files to ensure that the solution can be verified. The feature implementation requires LLMs to simultaneously possess code completion capabilities for new components and code editing abilities for other relevant parts in the code repository, providing a more comprehensive evaluation method of LLMs' automated software engineering capabilities. Experimental results show that LLMs perform significantly worse in the FEA-Bench, highlighting considerable challenges in such repository-level incremental code development.

As is often the case, Machine Learning (ML) techniques outperform traditional ones when allocating weights to different assets. The idea of this project "Portfolio Assets Allocation: A practical and scalable framework for Machine Learning Development" is to design a market neutral (long/short) portfolio of assets to be rebalanced periodically choosing different assets during every rebalance and evaluate different portfolio techniques such as: This article is the final project submitted by the author as a part of his coursework in the Executive Programme in Algorithmic Trading (EPAT) at QuantInsti. Do check our Projects page and have a look at what our students are building. Raimondo Marino is a professional freelance working as an Artificial intelligence Engineer for Italian Small and Medium Companies. Through AI applications, he comes up with end to end solutions (from Development to Production using cloud services) for different corporate functions within a company: Marketing, HR, Sales, Production, etc.

Google Developer is a platform that offers a wide range of tools, resources, and support for developers of all levels. Whether you are just starting out in your development journey or have been building applications for years, Google Developer has something to offer. One of the standout features of Google Developer is the abundance of APIs available. From Google Maps to Google Cloud, you can easily integrate these APIs into your projects to add powerful functionality. For example, the Google Maps API allows you to add interactive maps to your website or app, while the Google Cloud Platform allows you to build, deploy, and scale applications on Google's infrastructure.

Before we consider the five technologies that are set to transform Fintech, consider what Fintech is. Fintech is the synthesis of technology and finance and the harmonic combination of two of the largest industries into a single field. Naturally, its impact is enormous. Regarded as cutting-edge innovations a few years ago, now Fintech solutions are a daily reality. According to McKinsey, 80% of traditional financial institutions were exploring innovations in 2018.

MIT senior Rachel Chae and alumnus Sihao Huang '22 have been selected to join the 2023 class of Marshall Scholars and will begin graduate studies in the U.K. next fall. Funded by the British government, the Marshall Scholarship provides up to 50 scholarships for exceptional American students to pursue advanced study in any field at any university in the U.K. MIT's endorsed Marshall candidates are advised and supported by the distinguished fellowships team, led by Associate Dean Kim Benard in Career Advising and Professional Development. They are also mentored by the MIT Presidential Committee on Distinguished Fellowships, co-chaired by professors Will Broadhead and Tamar Schapiro. "Working with this year's Marshall applicants has been as rewarding and humbling as ever," says Broadhead. "These amazing students engage in a months-long exercise in critical introspection and personal growth, supported by the expert mentorship provided by Kim Benard and her team in the Distinguished Fellowships Office and by the dedicated faculty, staff, and graduate students who serve on the Distinguished Fellowships Committee. We on the committee have been inspired by all of this year's fellowship applicants and are especially pleased to congratulate Rachel and Sihao, whose wisdom, good humor, and future-minded optimism will serve them well as they take their richly deserved places in this year's class of Marshall Scholars."

Abstract: This paper focuses on the Matrix Factorization based Clustering (MFC) method which is one of the few closed-form algorithms for the subspace clustering algorithm. Despite being simple, closed-form, and computation-efficient, MFC can outperform the other sophisticated subspace clustering methods in many challenging scenarios. We reveal the connection between MFC and the Innovation Pursuit (iPursuit) algorithm which was shown to be able to outperform the other spectral clustering based methods with a notable margin especially when the span of clusters are close. A novel theoretical study is presented which sheds light on the key performance factors of both algorithms (MFC/iPursuit) and it is shown that both algorithms can be robust to notable intersections between the span of clusters. Importantly, in contrast to the theoretical guarantees of other algorithms which emphasized on the distance between the subspaces as the key performance factor and without making the innovation assumption, it is shown that the performance of MFC/iPursuit mainly depends on the distance between the innovative components of the clusters.

Abstract: Task planning is an important component of traditional robotics systems enabling robots to compose fine grained skills to perform more complex tasks. Recent work building systems for translating natural language to executable actions for task completion in simulated embodied agents is focused on directly predicting low level action sequences that would be expected to be directly executable by a physical robot. In this work, we instead focus on predicting a higher level plan representation for one such embodied task completion dataset TEACh, under the assumption that techniques for high-level plan prediction from natural language are expected to be more transferable to physical robot systems . We demonstrate that better plans can be predicted using multimodal context, and that plan prediction and plan execution modules are likely dependent on each other and hence it may not be ideal to fully decouple them. Further, we benchmark execution of oracle plans to quantify the scope for improvement in plan prediction models.

Sleep is an essential behavior to prevent the decrement of cognitive, motor, and emotional performance and various diseases. However, it is not easy to fall asleep when people want to sleep. There are various sleep-disturbing factors such as the COVID-19 situation, noise from outside, and light during the night. We aim to develop a personalized sleep induction system based on mental states using electroencephalogram and auditory stimulation. Our system analyzes users' mental states using an electroencephalogram and results of the Pittsburgh sleep quality index and Brunel mood scale. According to mental states, the system plays sleep induction sound among five auditory stimulation: white noise, repetitive beep sounds, rainy sound, binaural beat, and sham sound. Finally, the sleep-inducing system classified the sleep stage of participants with 94.7 percent and stopped auditory stimulation if participants showed non-rapid eye movement sleep. Our system makes 18 participants fall asleep among 20 participants.