Latest

Advancements in retrieving accessible information have evolved faster in the last few years compared to the decades since the internet’s creation. Search engines, like Google, have been the number one way to find relevant data. They have always relied on the user’s abilities to find the best...

The recent advancements in auto-regressive multimodal large language models (MLLMs) have demonstrated promising progress for vision-language tasks. While there exists a variety of studies investigating the processing of linguistic information within large language models, little is currently known...

Retrieval-Augmented Generation (RAG) has recently gained significant attention for its enhanced ability to integrate external knowledge sources into open-domain question answering (QA) tasks. However, it remains unclear how these models address fairness concerns, particularly with respect to...

This paper addresses the task of video question answering (videoQA) via a decomposed multi-stage, modular reasoning framework. Previous modular methods have shown promise with a single planning stage ungrounded in visual content. However, through a simple and effective baseline, we find that such...

Visual in-context learning (VICL), as a new paradigm in computer vision, allows the model to rapidly adapt to various tasks with only a handful of prompts and examples. While effective, the existing VICL paradigm exhibits poor generalizability under distribution shifts. In this work, we propose...

Large language models (LLMs) have demonstrated remarkable advancements in language understanding and generation. Building on the success of text-based LLMs, recent research has adapted these models to use speech embeddings for prompting, resulting in Speech-LLM models that exhibit strong performance...

Large language models (LLMs) have been effectively used for many computer vision tasks, including image classification. In this paper, we present a simple yet effective approach for zero-shot image classification using multimodal LLMs. Using multimodal LLMs, we generate comprehensive textual...

Language models are aligned to the collective voice of many, resulting in generic outputs that do not align with specific users’ styles. In this work, we present Trial-Error-Explain In-Context Learning} (ITCL), a tuning-free method that personalizes language models for text generation tasks with...

Accurate greenhouse gas (GHG) emission reporting is critical for governments, businesses, and investors. However, adoption remains limited particularly among small and medium enterprises due to high implementation costs, fragmented emission factor databases, and a lack of robust sector classification...

Large language models accumulate vast knowledge during pre-training, yet the dynamics governing this acquisition remain poorly understood. This work investigates the learning dynamics of language models on a synthetic factual recall task, uncovering three key findings: First, language models learn in...

Accurate camera calibration is a fundamental task for 3D perception, especially when dealing with real-world, in-the-wild environments where complex optical distortions are common. Existing methods often rely on pre-rectified images or calibration patterns, which limits their applicability and...

Modern AI systems such as self-driving cars and game-playing agents achieve superhuman performance, but often lack human-like generalization, interpretability, and inter-operability with human users. Inspired by the rich interactions between language and decision-making in humans, we introduce Policy...

Customized text-to-video generation aims to produce high-quality videos that incorporate user-specified subject identities or motion patterns. However, existing methods mainly focus on personalizing a single concept, either subject identity or motion pattern, limiting their effectiveness for multiple...

Open-vocabulary semantic segmentation models associate vision and text to label pixels from an undefined set of classes using textual queries, providing versatile performance on novel datasets. However, large shifts between training and test domains degrade their performance, requiring fine-tuning...

We present StyleMotif, a novel Stylized Motion Latent Diffusion model, generating motion conditioned on both content and style from multiple modalities. Unlike existing approaches that either focus on generating diverse motion content or transferring style from sequences, StyleMotif seamlessly...

We introduce LOCORE, Long-Context Re-ranker, a model that takes as input local descriptors corresponding to an image query and a list of gallery images and outputs similarity scores between the query and each gallery image. This model is used for image retrieval, where typically a first ranking is...

Recent advancements in multimodal large language models (MLLMs) have focused on integrating multiple modalities, yet their ability to simultaneously process and reason across different inputs remains underexplored. We introduce OmniBench, a novel benchmark designed to evaluate models’ ability to...

High-quality view synthesis relies on volume rendering, splatting, or surface rendering. While surface rendering is typically the fastest, it struggles to accurately model fuzzy geometry like hair. In turn, alpha-blending techniques excel at representing fuzzy materials but require an unbounded...

There exist three approaches for multilingual and crosslingual automatic speech recognition (MCL-ASR) - supervised pretraining with phonetic or graphemic transcription, and self-supervised pretraining. We find that pretraining with phonetic supervision has been underappreciated so far for MCL-ASR,...

Large Language Models (LLMs) have shown remarkable capabilities across various domains, but their potential for solving combinatorial optimization problems remains largely unexplored. In this paper, we investigate the applicability of LLMs to the Job Shop Scheduling Problem (JSSP), a classic...

An AI overhaul may be on the horizon for Apple’s Health app. In the Power On newsletter, Bloomberg’s Mark Gurman reports that Apple is working on a much more comprehensive version of its Health app under the code name Project Mulberry, with plans to integrate an AI agent that would somewhat...

Non-paying Gemini users can now play around with Google’s newest model, the experimental version of Gemini 2.5 Pro. The company announced this weekend that it’s making Gemini 2.5 Pro (experimental) free for everyone to use, albeit with tighter rate limits for non-subscribers. Google introduced Gemini...

The Amazon Spring Sale is here, bringing a slew of discounts on household essentials, fashion, outdoor gear and even furniture. We at Engadget are focused on tech, and while the selection this time around isn’t quite as good as that of Amazon Prime Day in July, there are some decent discounts to be...

Amazon kicked off its latest Spring Sale last week, and while the event mostly focuses on things like apparel and outdoor goods, it still includes a few noteworthy deals on electronics. We’re rounding up the best of those across Engadget, but if you’re on a tighter budget, below we’ve dug up a few...

We might not have personal robot butlers that can clean our entire home, but when it comes to automatically cleaning floors, the future has been delivered. And thanks to Amazon’s Spring Sale, a number of models are currently discounted. We’ve tested dozens of robot vacuums to produce two guides on...

Amazon’s latest Spring Sale has arrived. While the event isn’t as heavy on tech as Black Friday or Prime Day, we’ve still managed to find a few noteworthy discounts on Apple devices. If you’re looking for a new iPad, Apple Watch or set of AirPods, we’ve rounded up the best Apple deals going on right...

The Amazon Big Spring Sale isn’t as huge as, say, the company’s (two) Prime Day sales, but there are still plenty of deals to be found — particularly on goods for your home. While Engadget doesn’t test sheets and pillows, we do put our review skills to work on kitchen tech. We’ve combed through...

An excerpt from the upcoming book “The Optimist: Sam Altman, OpenAI, and the Race to Invent the Future” offers new details about why OpenAI’s board briefly fired CEO Sam Altman back in 2023. Written by Wall Street Journal reporter Keach Hagey, the book claims the nonprofit’s board members became...

Amazon announced this week that Jennifer Salke is stepping down as the head of Amazon MGM Studios — a move that could indicate high-level dissatisfaction with the company’s streaming strategy over the past few years. To be clear, Salke’s boss Mike Hopkins (the head of Amazon MGM Studios and Prime...

Welcome back to Week in Review! Elon Musk says that xAI bought X in an all-stock deal; a16z-backed 11x faked some customers; the Lumon Terminal Pro pops up on Apple’s website; and much more. Let’s dig in! xAI buys X: X owner Elon Musk announced on X on Friday that xAI acquired X in an […]

United Football League’s partnership with Salesforce brings fans closer to the game with instant customer service and the first AI agent-powered broadcast.

In this issue, we examine a new conversation segmentation method that delivers more coherent and personalized agent conversation, and we review efforts to improve MLLMs’ understanding of geologic maps. Check out the latest research and other updates. The post Research Focus: Week of March 24, 2025...

Today, Google announced new short-term precipitation forecasts are now available across Africa in Search. This expands upon our work to use machine learning to tackle th…

All Google Workspace for Education users have added data protection in the Gemini app, meaning their chats with Gemini are not human reviewed or used to train AI models.…

Training separate models on different datasets and then merging them reduces computational costs by as much as 91%.

Our team at Salesforce Research introduces Text2Data, an innovative framework specifically designed to generate high-quality, controllable data from limited textual input.

Since the release of ChatGPT in November 2022, the capabilities of large language models (LLMs) have surged, and the number of available models has grown...

Large language models (LLMs) often struggle with accuracy when handling domain-specific questions, especially those requiring multi-hop reasoning or access to...

AI is rapidly transforming industries, empowering businesses to enhance customer experiences, improve efficiency, and make smarter decisions. But an essential question arises: How can we ensure that AI is creating accurate and grounded…

Our latest Workspace drop is here and includes some helpful Gemini updates for Google Meet business users. Here are few new things to try:The “Take notes for me” feature…