Latest

Despite advancements in multimodal large language models (MLLMs), current approaches struggle in medium-to-long video understanding due to frame and context length limitations. As a result, these models often depend on frame sampling, which risks missing key information over time and lacks...

Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for empowering large language model (LLM) applications. Compared with the supervised training process of LLMs, the RLHF training process is much more sophisticated, requiring a diverse range of computation workloads with...

Geometry problem-solving (GPS), a challenging task requiring both visual comprehension and symbolic reasoning, effectively measures the reasoning capabilities of multimodal large language models (MLLMs). Humans exhibit strong reasoning ability in this task through accurate identification and adaptive...

Conversational Product Search ( CPS ) systems interact with users via natural language to offer personalized and context-aware product lists. However, most existing research on CPS is limited to simulated conversations, due to the lack of a real CPS dataset driven by human-like language. Moreover,...

Sparse attention offers a promising strategy to extend long-context capabilities in Transformer LLMs, yet its viability, its efficiency-accuracy trade-offs, and systematic scaling studies remain unexplored. To address this gap, we perform a careful comparison of training-free sparse attention methods...

We present GameNGen, the first game engine powered entirely by a neural model that also enables real-time interaction with a complex environment over long trajectories at high quality. When trained on the classic game DOOM, GameNGen extracts gameplay and uses it to generate a playable environment...

Researchers have recently suggested that models share common representations. In our work, we find that token embeddings of language models exhibit common geometric structure. First, we find ``global’’ similarities: token embeddings often share similar relative orientations. Next, we characterize...

Human activity recognition (HAR) on smartglasses has various use cases, including health/fitness tracking and input for context-aware AI assistants. However, current approaches for egocentric activity recognition suffer from low performance or are resource-intensive. In this work, we introduce a...

This study addresses the critical challenge of hallucination mitigation in Large Vision-Language Models (LVLMs) for Visual Question Answering (VQA) tasks through a Split Conformal Prediction (SCP) framework. While LVLMs excel in multi-modal reasoning, their outputs often exhibit hallucinated content...

We present DyMU, an efficient, training-free framework that dynamically reduces the computational burden of vision-language models (VLMs) while maintaining high task performance. Our approach comprises two key components. First, Dynamic Token Merging (DToMe) reduces the number of visual token...

Daily Activity Recordings for Artificial Intelligence (DARai, pronounced "Dahr-ree") is a multimodal, hierarchically annotated dataset constructed to understand human activities in real-world settings. DARai consists of continuous scripted and unscripted recordings of 50 participants in 10 different...

As false information continues to proliferate across social media platforms, effective rumor detection has emerged as a pressing challenge in natural language processing. This paper proposes RAGAT-Mind, a multi-granular modeling approach for Chinese rumor detection, built upon the MindSpore deep...

Scalable general-purpose representations of the built environment are crucial for geospatial artificial intelligence applications. This paper introduces S2Vec, a novel self-supervised framework for learning such geospatial embeddings. S2Vec uses the S2 Geometry library to partition large areas into...

The demand for high-fidelity test data is paramount in industrial settings where access to production data is largely restricted. Traditional data generation methods often fall short, struggling with low-fidelity and the ability to model complex data structures and semantic relationships that are...

Large language models (LLMs) are a transformational capability at the frontier of artificial intelligence and machine learning that can support decision-makers in addressing pressing societal challenges such as extreme natural hazard events. As generalized models, LLMs often struggle to provide...

Extremely low-resource (XLR) languages lack substantial corpora for training NLP models, motivating the use of all available resources such as dictionaries and grammar books. Machine Translation from One Book (Tanzer et al., 2024) suggests that prompting long-context LLMs with one grammar book...

Accurate interpretation of lab results is crucial in clinical medicine, yet most patient portals use universal normal ranges, ignoring conditional factors like age and gender. This study introduces Lab-AI, an interactive system that offers personalized normal ranges using retrieval-augmented...

Abstractive dialogue summarization is the task of distilling conversations into informative and concise summaries. Although reviews have been conducted on this topic, there is a lack of comprehensive work detailing the challenges of dialogue summarization, unifying the differing understanding of the...

We present our shared task on text-based emotion detection, covering more than 30 languages from seven distinct language families. These languages are predominantly low-resource and are spoken across various continents. The data instances are multi-labeled with six emotional classes, with additional...

Large Language Models (LLMs) have demonstrated strong reasoning capabilities in solving complex problems. However, current approaches primarily enhance reasoning through the elaboration of thoughts while neglecting the diversity of reasoning types. LLMs typically employ deductive reasoning,...

Microsoft is also adding ’Researcher’ and ’Analyst’ agents powered by OpenAI’s latest reasoning models.

ChatGPT, OpenAI’s text-generating AI chatbot, has taken the world by storm since its launch in November 2022. What started as a tool to supercharge productivity through writing essays and code with short text prompts has evolved into a behemoth with 300 million weekly active users. 2024 was a big...

Kai Chen, a Canadian AI researcher working at OpenAI who’s lived in the U.S. for 12 years, was denied a green card, according to Noam Brown, a leading research scientist at the company. In a post on X, Brown said that Chen learned of the decision Friday and must soon leave the country. “It’s deeply...

Google started testing AI-summarized results in Google Search, AI Overviews, two years ago, and continues to expand the feature to new regions and languages. By the company’s estimation, it’s been a big success. AI Overviews is now used by more than 1.5 billion users monthly across over 100...

Google’s program is available in 40 countries.

Chinese startup Manus AI, which works on building tools related to AI agents, has picked up $75 million in a funding round led by Benchmark at a roughly $500 million valuation, according to Bloomberg. The company will use the money to expand to new markets, including the U.S., Japan, and the Middle...

Ascoria’s Tim Cole discusses how technology is shaking up the taxation and compliance industry, and why he’s excited about agentic AI. Read more: How digital transformation is ‘rewiring’ the tax industry

Google’s cloud revenues jumped by 28pc in Q1, a key focus for the company as its adtech and search empire comes under growing antitrust scrutiny. Read more: Focus on cloud revenues as Alphabet earnings announced

GPUs from AMD and Nvidia probably helped, too.

Apple plans to relocate the assembly of all iPhones destined for the US market from China to India as early as next year.

Explore the different types of AI agents — sales, marketing, service, and commerce — and see how they can help you enhance productivity.

AI agents need a structured metadata map to make sense of how information is connected so they can generate meaningful outcomes.

Google leaders as well as policymakers, industry partners and civil society groups stand on a stage with the event title behind them

Read Google and Alphabet CEO Sundar Pichai’s remarks from the Q1 2025 earnings call.

Designing for the age of intelligent systems means rethinking how we structure content, interpret language, and support user intent.

Our latest Google Workspace feature drop is here with new tools to help you collaborate, stay on top of your to-do list and do your best work. Plus, this latest update h…

Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Integer vitae fermentum sem. Duis cursus erat in luctus euismod.

Illustration of Earth in dark and light blue, with four white clouds, surrounded by blue dots and lines

3 illustrations, woman and daughter doing homework; woman icing a cake; man on his phone.

AI could one day help neuroscientists predict activity in the human brain, just like language models can predict the next word in a sentence.After more than a decade of …