Results for “model”

arXiv (research papers) Bolin Zhang

SCOPE: A Physical Execution Latency-Guided Quantum Circuit Compilation Framework

Quantum computing requires compilers to translate algorithms represented in the quantum circuit model into unitary operations supported by physical devices. Limited coherence time makes physical execution latency an important compilation objective. Gate fusion can shorten pulses, but gate count and depth do not capture fused targets' control demands. Evaluating every candidate with quantum optimal control (QOC) would embed pulse optimization in structural search. We present…

Read at arXiv (research papers)

arXiv (research papers) Sahaya Jestus Lazer SI

Evaluating Authorization and Trust Boundaries in Agentic Orchestration Frameworks

Agentic AI systems extend large language models with tools, memory, and multi-agent collaboration, introducing security risks at orchestration boundaries. These risks depend on how identities are represented, operations are authorized, and instructions or information propagate between components. We evaluate LangChain, LangGraph, AutoGen, and OpenClaw through twelve experiments mapped to the OWASP Top 10 for Agentic Applications using a common LLM (gpt-4o-mini). The…

Read at arXiv (research papers)

arXiv (research papers) Haoyun Tang

ThinkWAM: Rethinking Latent Futures for Action Refinement

World-action models (WAMs) couple action generation with predictions of future states. An initial action chunk may be flawed. A future prediction that is not conditioned on the chunk may fail to reflect the changes that the proposed actions would induce in visual representations, limiting its usefulness for refinement before execution. We introduce ThinkWAM, a latent WAM that establishes bidirectional feedback before execution through a proposal-rethink-refine loop: the…

Read at arXiv (research papers)

arXiv (research papers) Lucas Schulze

BoBDeLaN: Bounded-Bandwidth Context Adaptation for Deep Lagrangian Robot Control

Learned physics-based models provide structure for robot control, but uncertain payloads and unmodeled dynamics limit their predictive accuracy. Context-aware Deep Lagrangian Networks enable adaptation through a latent context inferred from measurement histories while retaining Lagrangian structure for each fixed context. However, this structure alone does not constrain sensitivity to measurement disturbances. We introduce BoBDeLaN, which combines a contracting, gain-bounded…

Read at arXiv (research papers)

9to5Mac Ryan Christoffel Gadgets

Next year’s iPhone Pro redesign gets new update via leaker

Next year is the iPhone’s 20th anniversary, and Apple reportedly has a big redesign coming to its Pro models. Here’s what to expect per the latest leak.

Read at 9to5Mac

arXiv (research papers) Hua Zhang

Linglong Agent Service: A Runtime for Centralized Agent Reasoning and Distributed Execution

LLM agents can use remote tools and cloud sandboxes, but integrating reasoning, session state, and heterogeneous execution sites remains a runtime design problem. We present Linglong Agent Service (agentsvc), a self-hosted HTTP service and embeddable Python runtime that treats execution site as a configurable dimension alongside model, tools, and context. A parent service retains inference orchestration, tool descriptions, and conversation history; each inference request…

Read at arXiv (research papers)

The Hacker News Security

Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

Google on Monday revealed that adversarial use of artificial intelligence (AI) by threat actors has gone beyond manual prompting to deploying autonomous multi-agent offensive frameworks like ARTEX, CAIRN, CyberStrikeAI, and PentAGI to execute multi-stage intrusions at greater operational velocity. In tandem, threat actors are leveraging the power and capabilities of large language models (LLMs)

Read at The Hacker News

Google (The Keyword)

Ask Maps is coming to Europe.

Ask Maps is the biggest update to Google Maps in over a decade. Powered by the world’s freshest map and Gemini models, it answers complex, real-world questions a map cou…

Read at Google (The Keyword)

MacRumors Mitchel Broussard Gadgets

Apple Watch SE 3 Black Friday Deals Arrive With $50 Off Every Model

Following in the footsteps of this weekend's big discounts on the Apple Watch Series 12, Amazon today has all-time low prices across a number of Apple Watch SE 3 devices. Prices start at just $199.00 for the 40mm GPS Apple Watch SE 3, matching the previous record low price on this model and marking the first time we've tracked a discount on the SE 3 this Black Friday season. Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we…

Read at MacRumors

Digital Trends Shimul Sood

AI models will change their response if you warn that it’s being monitored, but not necessarily for good

A new study involving 10 AI models suggests that their willingness to follow instructions can change depending on whether they believe their responses are being monitored. The findings raise questions about how reliably AI systems behave outside controlled testing environments.

Read at Digital Trends

arXiv (research papers) Ke Sun

ForceHarness: Turning Force into Action and Physics for Frozen Frontier Models

Frontier multimodal models can now control robots zero-shot from images, yet many manipulation tasks depend on contact forces that images cannot show. Bringing force to an API-only frozen model raises two challenges. First, contact evolves much faster than an API model can decide, so the system must close the loop on force beneath the model. Second, a frozen model cannot interpret raw force streams, so force must be returned in a form the model can reason about. Therefore…

Read at arXiv (research papers)

MacRumors Hartley Charlton Gadgets

'iPhone 20 Pro' Rumored to Be Based on iPhone Air

Next year's "iPhone 20 Pro" models will use the same glass design as the iPhone Air, but with more pronounced curvature along all four sides, according to the leaker known as "Fixed Focus Digital." In a new post on Weibo, the leaker said the device would use the same glass material as the ‌iPhone Air‌, adjusted to have greater curvature and produce a four-sided curved effect. The accompanying hashtag refers specifically to an "iPhone 20 Pro" with a four-sided…

Read at MacRumors

arXiv (research papers) Guangzhi Sun Security

AudioTrace: File-Level Membership Detection for Audio Large Language Models

Significant privacy and copyright concerns are raised by the use of publicly available audio recordings to train audio large language models (LLMs), yet effective auditing methods remain limited. Existing membership inference attacks typically assume knowledge of complete training examples, including their associated textual supervision. In practice, however, audio recordings may be paired with privately generated transcripts, captions, or question-answer pairs that are…

Read at arXiv (research papers)

Digital Trends Anvinraj Valiyathara

Qwen4-Max is China’s newest flagship AI model, and it makes GPT and Claude look expensive

Alibaba's Qwen4-Max costs $1.50/$4.50 per million tokens, cuts task costs by 62%, and was trained to avoid offensive hacking skills.

Read at Digital Trends

ZDNET AI Charlie Osborne SI

AI finds zero-day flaws for just $4 now – even mediocre hackers can cash in

Rapidly advancing AI models are making it cheaper and easier to uncover critical vulnerabilities, but it’s a double-edged sword – and a gift to hackers.

Read at ZDNET AI

MIT Technology Review AI MIT Technology Review Insights SI

Building AI that can work across the enterprise

For years, enterprises have approached AI as a series of discrete experiments. Now, as AI agents begin to move into the workflows that run the business, that model is starting to look inadequate. Organizations moving from isolated pilots toward AI agents and autonomous workflows are confronting a deeper challenge: rethinking how data, business processes, governance,…

Read at MIT Technology Review AI

Analytics India Magazine Mukundan Sivaraj SI

Alibaba Launches Qwen4-Max AI Model With Focus on Security

The 2.4-trillion-parameter model includes a 1-million-token context window, native vision capabilities and security improvements across coding and other tasks.

Read at Analytics India Magazine

ReliefWeb World Food Programme

WFP Armenia Country Brief, November 2026

Country: Armenia Source: World Food Programme Please refer to the attached file. KEY HIGHLIGHTS Armenia positioned its National School Meals Programme as a flagship, government-owned model at CBD COP17, demonstrating how school meals can simultaneously advance nutrition, education, local food systems, biodiversity conservation, climate resilience, and human capital development. WFP continued scaling the Transformative Home-Grown School Feeding model and nutrition-sensitive…

Read at ReliefWeb

arXiv (research papers) Mika Facchini Technical University of Munich Tech policy

Auditing a Moving Target: Evidence Anchoring and the Expiry of Fairness Audits of Provider-Served Language Models

Organizations increasingly use large language models (LLMs) for decisions about people, such as hiring, investment and employee evaluation, and rely on them over long periods of time. Fairness audits of such systems, however, are conducted only once. We examine how the biases of LLMs change over time. In a venture screening setting, models from five providers decided whether and how much to invest in startups whose founders differed only in gender, origin and university. We…

Read at arXiv (research papers)

Tech.eu John Reynolds Startups

AI inference startup Daisytuner raises €6M seed

A German startup whose technology tunes AI models for different chips during inference has raised €6 million in seed funding, it said today.Daisytuner has raised the funds in a round co-led by Elaia a...

Read at Tech.eu

arXiv (research papers) Zhiming Chi Security

Scenario-Based Risk Certificates for Multi-Property Vulnerability Discovery in Parametric MDPs

Safety-critical systems such as network protocols and distributed algorithms depend on environmental parameters whose exact values are uncertain. Parametric Markov decision processes (pMDPs) model such systems by making transition probabilities symbolic functions of the parameters. For a single quantitative requirement, parametric verification asks which parameter valuations satisfy the requirement. In practice, however, engineers often monitor several safety and performance…

Read at arXiv (research papers)

arXiv (research papers) Ulysse Pavloff Security

Optimal Dynamic Fees for L2s: A Vault Approach

Layer 2 (L2) blockchains incur costs when posting data to Layer 1 (L1) and charge users fees intended to recover those costs. We model this mechanism as a *vault*: user fees increase its balance, L1 posting costs reduce it, and surplus and deficit accumulate over time. Forecast-based fees aim to recover posting costs, but forecast errors accumulate in the vault. Returning the vault balance to a target requires feedback on the accumulated error. Treating the vault as a…

Read at arXiv (research papers)

VietnamPlus (Vietnamese state media) Geopolitics

Azerbaijani President to pay official visit to Vietnam

The official visit from November 20 – 21 will be made at the invitation of General Secretary of the Communist Party of Vietnam Central Committee and President To Lam. Vietnam introduces community-based tourism models at Azerbaijan conference Top leader receives Azerbaijani presidential envoy Vietnamese Vice President holds talks with Azerbaijani leaders

Read at VietnamPlus (Vietnamese state media)

BioPharma Dive

The infrastructure decisions that will enable tomorrow’s biggest breakthroughs

Faster drug discovery depends on more than the model. What drug developers need to know about AI.

Read at BioPharma Dive

arXiv (research papers) Francesco Cerasuolo Security

The Cybersecurity Dataset Observatory: Which Data for AI in Cybersecurity?

AI is increasingly used for cybersecurity tasks such as intrusion and malware detection, in enterprise, industrial, vehicular and mobile networks. Its effectiveness depends on data: models need representative network traffic, both to be trained and to be evaluated realistically. Public datasets of cybersecurity network traffic, however, are scattered across laboratory pages, repositories and aggregators, and are continuously released, corrected and withdrawn: 92 of the 234…

Read at arXiv (research papers)

arXiv (research papers) Wenrui Zhang

QuarkDB: A Hybrid Database System for Quantum Data Management

Emerging quantum applications like quantum communication generate large volumes of high-dimensional quantum data, requiring efficient data analysis and storage. However, classical database systems often struggle to meet requirements such as quantum processing or version management. This paper presents QuarkDB designed for efficient quantum data management. QuarkDB establishes a unified object model for relational and quantum data and defines the syntax of Quantum Structured…

Read at arXiv (research papers)

MarkTechPost Asif Razzaq SI

Alibaba Releases Qwen4-Max: 2.4T MoE, 71B Active, 1M Context

Alibaba released Qwen4-Max, a proprietary 2.4T-parameter MoE model with 71B active parameters, native vision and a 1M-token context. It costs $1.5/M input tokens and cuts Work360 cost per task by 62%.

Read at MarkTechPost

OpenAI News SI

How Revolut raises ambition without lowering the bar

Discover how Revolut uses OpenAI models and Codex to cut coding time by 30–60%, scale financial crime investigations, and innovate without compromising security.

Read at OpenAI News

arXiv (research papers) Zhe Dong

Building Fully On-Device GUI Agents with Mobile NPUs

Mobile GUI agents are increasingly used to automate everyday smartphone tasks, but most existing systems still rely on cloud-hosted models or remote inference, introducing repeated communication latency, privacy exposure, and dependence on network availability. Modern smartphones integrate neural processing units (NPUs) that provide an attractive substrate for fully on-device LLM inference. However, directly deploying conventional GUI-agent pipelines on mobile NPUs is…

Read at arXiv (research papers)

ReliefWeb International Committee of the Red Cross

World: From Principles to Practice – AI in Humanitarian Organizations (AI Alignment Guideline, 2026)

Country: World Source: International Committee of the Red Cross Please refer to the attached file. This practical guide helps humanitarian organization integrate the principles of humanity, impartiality, neutrality and independence into the design, deployment and use of AI systems, with a focus on large language models and agentic AI. Building on existing AI policy and ethical guidelines, it helps humanitarian practitioners explore, assess and improve the alignment of AI…

Read at ReliefWeb

Older stories