stp2y

34355 Posts
Active offline policy selection

Active offline policy selection

Reinforcement learning (RL) has made tremendous progress in recent years towards addressing real-life problems – and offline RL made it even more practical. Instead of direct interactions with the environment, we can now train many algorithms from a single pre-recorded dataset. However, we lose the practical advantages in data-efficiency of offline RL when we evaluate the policies at hand.For example, when training robotic manipulators the robot resources are usually limited, and training many policies by offline RL on a single dataset gives us a large data-efficiency advantage compared to online RL. Evaluating each policy is an expensive process, which requires…
Read More
Efficient 3D-Aware Facial Image Editing via Attribute-Specific Prompt Learning

Efficient 3D-Aware Facial Image Editing via Attribute-Specific Prompt Learning

arXiv:2406.04413v1 Announce Type: new Abstract: Drawing upon StyleGAN's expressivity and disentangled latent space, existing 2D approaches employ textual prompting to edit facial images with different attributes. In contrast, 3D-aware approaches that generate faces at different target poses require attribute-specific classifiers, learning separate model weights for each attribute, and are not scalable for novel attributes. In this work, we propose an efficient, plug-and-play, 3D-aware face editing framework based on attribute-specific prompt learning, enabling the generation of facial images with controllable attributes across various target poses. To this end, we introduce a text-driven learnable style token-based latent attribute editor (LAE). The LAE…
Read More
Large Language Model Confidence Estimation via Black-Box Access

Large Language Model Confidence Estimation via Black-Box Access

arXiv:2406.04370v1 Announce Type: new Abstract: Estimating uncertainty or confidence in the responses of a model can be significant in evaluating trust not only in the responses, but also in the model as a whole. In this paper, we explore the problem of estimating confidence for responses of large language models (LLMs) with simply black-box or query access to them. We propose a simple and extensible framework where, we engineer novel features and train a (interpretable) model (viz. logistic regression) on these features to estimate the confidence. We empirically demonstrate that our simple framework is effective in estimating confidence of flan-ul2,…
Read More
WWDC 2024: How to watch Apple’s keynote on iOS 18, AI and more

WWDC 2024: How to watch Apple’s keynote on iOS 18, AI and more

Apple’s Worldwide Developers Conference (WWDC) keynote is imminent. The festivities kick off later today — Monday, June 10 at 1PM ET. The keynote address is available to the public and you can watch it via Apple’s event website or on the company’s YouTube channel. And if you don't want to click away, the latter feed is embedded directly below.This is WWDC, so it’ll be a software-focused event. Expect that Apple will showcase updates across its full panoply of operating systems, including iOS 18 and iPadOS 18, as well as watchOS, macOS and even visionOS, which is the operating system behind…
Read More
AI Tools Are Secretly Training on Real Images of Children

AI Tools Are Secretly Training on Real Images of Children

Over 170 images and personal details of children from Brazil have been scraped by an open-source dataset without their knowledge or consent, and used to train AI, claims a new report from Human Rights Watch released Monday.The images have been scraped from content posted as recently as 2023 and as far back as the mid-1990s, according to the report, long before any internet user might anticipate that their content might be used to train AI. Human Rights Watch claims that personal details of these children, alongside links to their photographs, were included in LAION-5B, a dataset that has been a…
Read More
Intern day 1: My first steps with real-time data software | Deephaven

Intern day 1: My first steps with real-time data software | Deephaven

Hi, I'm Josh, one of Deephaven's newest interns. I'm a computer science major and looking forward to learning about data science in the real world while also expanding my software abilities. Although I'd heard great things about Deephaven's query engine, I had never used it before. As a part of my job, I was expected to learn it from scratch within a week. Learning any new software can be intimidating, especially when its tutorials lack the information you need. To my surprise, it only took two days to get comfortable enough with the IDE and the Deephaven Query Language to…
Read More
Russia is attempting to protect a vital bridge and supply line with a set of barges, UK intelligence says

Russia is attempting to protect a vital bridge and supply line with a set of barges, UK intelligence says

Russia is shoring up the defenses for the Kerch Bridge with barges, the UK's defense ministry said on Saturday."Analysis of imagery has identified the installation of eight barges on the southern side of the Kerch Bridge," the UK's defense ministry wrote in its intelligence dispatch."These barges were placed by Russian forces in an attempt to defend the bridge and shipping channel, reducing the angles of approach for Ukrainian Unmanned Surface Vehicles (USVs)," the ministry said.According to the ministry, the barges were installed between May 10 and May 22. This isn't the first time barges have been placed near the bridge,…
Read More
Why Has Predicting Downstream Capabilities of Frontier AI Models with Scale Remained Elusive?

Why Has Predicting Downstream Capabilities of Frontier AI Models with Scale Remained Elusive?

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. Source link lol
Read More
No widgets found. Go to Widget page and add the widget in Offcanvas Sidebar Widget Area.