Tech + AI + Science NewsTech + AI + Science News
11 item(s)
Tech + AI + Science News
Remove AI Watermarks
Hacker News
Remove AI Watermarks new
Hacker News · 2026-05-19
Comments
Tech + AI + Science News
Tesla's lithium refinery discharges 231,000 gallons of
Hacker News
Tesla's lithium refinery discharges 231,000 gallons of polluted wastewater a day new
Hacker News · 2026-05-19
Comments
Tech + AI + Science News
OpenAI Adopts Google's SynthID Watermark for AI
Hacker News
OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool new
Hacker News · 2026-05-19
Comments
Tech + AI + Science News
Dumb Ways for an Open Source Project
Hacker News
Dumb Ways for an Open Source Project to Die new
Hacker News · 2026-05-19
Comments
Tech + AI + Science News
OpenAI and Dell partner to bring Codex
OpenAI News
OpenAI and Dell partner to bring Codex to hybrid and on-premise enterprise environments
OpenAI News · 2026-05-18
OpenAI and Dell partner to bring Codex to hybrid and on-premise environments, helping enterprises deploy AI coding agents securely across data and workflows.
Tech + AI + Science News
EU AI Act Update: Timeline Relief, Targeted
Brave News
EU AI Act Update: Timeline Relief, Targeted Simplification, and New Prohibitions | Inside Privacy new
Brave News
On 7 May 2026, negotiators from the Council of the European Union, the European Parliament, and the European Commission reached a provisional agreement on
Tech + AI + Science News
AI Model Rankings May 2026: Top LLMs
Brave News
AI Model Rankings May 2026: Top LLMs Ranked by Coding, Reasoning & Cost new
Brave News
A May 2026 snapshot of the top large language models ranked across three axes that actually matter: SWE-bench coding, GPQA Diamond reasoning, and real price per million tokens.
Tech + AI + Science News
How AI is transforming network incident response
Brave News
How AI is transforming network incident response (and where it still falls short) | Network World new
Brave News
AI is great at cutting down alert noise and spotting anomalies, but it can't fix your network if it's blind to 30% of the data paths in the first place.
Tech + AI + Science News
The Ed-Tech Backlash Is Here. What It
Brave News
The Ed-Tech Backlash Is Here. What It Means for Schools
Brave News
Most educators—74%—say their ... technology due to pushback or complaints from parents, according to the EdWeek Research Center survey. The percentage of districts dialing back tech use in school could rise as generative AI increasingly becomes integral to ed-tech tools, and educators, parents, and students see the downsides of too much AI use. Surveys from multiple organizations so far show that while parents want their children to learn how to use ...
Tech + AI + Science News
Take your local GitHub sessions anywhere
GitHub Blog
Take your local GitHub sessions anywhere
GitHub Blog · 2026-05-18
Kick off work in VS Code or the CLI, finish it from your phone. Remote control for GitHub Copilot sessions is now generally available on github.com and GitHub Mobile. The post Take your local GitHub sessions anywhere appeared first on The GitHub Blog .
Tech + AI + Science News
Forget electrons, this breakthrough uses light-matter particles
ScienceDaily — Artificial Intelligence
Forget electrons, this breakthrough uses light-matter particles to power AI new
ScienceDaily — Artificial Intelligence · 2026-05-19
Researchers at Penn have created a hybrid light-matter particle that could dramatically speed up AI computing while using far less energy. The breakthrough may help replace some electronic computing processes with ultra-efficient light-based technology.
Research (arXiv)
10 item(s)
Research (arXiv)
Can These Views Be One Scene? Evaluating
arXiv
Can These Views Be One Scene? Evaluating Multiview 3D Consistency when 3D Foundation Models Hallucinate new
arXiv · 2026-05-18
Multiview 3D evaluation assumes that the images being scored are observations of one static 3D scene. This assumption can fail in NVS and sparse-view reconstruction: inputs or generated outputs may contain artifacts, outlier frames, repeated views, or noise, yet still receive high 3D consistency scores. Existing reference-based metrics require ground truth, while ground-truth-free metrics such as MEt3R depend on learned reconstruction backbones whose failure modes are poorly characterized. We study this reliability problem by comparing neural reconstruction priors with classical geometric verification. We introduce \benchmark, a controlled robustness benchmark for multiview 3D consistency, and a parametric family that decomposes neural metrics into backbone, residual, and aggregation components. This family recovers MEt3R and yields variants up to $3\times$ more robust. Our analysis show
Research (arXiv)
DashAttention: Differentiable and Adaptive Sparse Hierarchical…
arXiv
DashAttention: Differentiable and Adaptive Sparse Hierarchical Attention new
arXiv · 2026-05-18
Current hierarchical attention methods, such as NSA and InfLLMv2, select the top-k relevant key-value (KV) blocks based on coarse attention scores and subsequently apply fine-grained softmax attention on the selected tokens. However, the top-k operation assumes the number of relevant tokens for any query is fixed and it precludes the gradient flow between the sparse and dense stages. In this work, we propose DashAttention (Differentiable and Adaptive Sparse Hierarchical Attention), which leverages the adaptively sparse $α$-entmax transformation to select a variable number of blocks according to the current query in the first stage. This in turn provides a prior for the second-stage softmax attention, keeping the entire hierarchy fully differentiable. Contrary to other hierarchical attention methods, we show that DashAttention is non-dispersive, translating to better long-context modeling
Research (arXiv)
A Readiness-Driven Runtime for Pipeline-Parallel Training under
arXiv
A Readiness-Driven Runtime for Pipeline-Parallel Training under Runtime Variability new
arXiv · 2026-05-18
Pipeline parallelism is a key technique for scaling large-model training, but modern workloads exhibit runtime variability in computation and communication. Existing pipeline systems typically consume static, profiled, or adaptively generated schedules as pre-committed execution orders. When realized task readiness diverges from the pre-committed order, stages may wait for not-yet-ready work even though other executable work is available, creating stage misalignment, idle bubbles, and reduced utilization. We present Runtime-Readiness-First Pipeline (RRFP), a readiness-driven runtime for pipeline-parallel training. RRFP changes how schedules are consumed at runtime: instead of treating a schedule as a sequence that stages must wait to follow, it treats the schedule as a non-binding hint order for ranking currently ready work. To support this model, RRFP combines message-driven asynchronou
Research (arXiv)
WavFlow: Audio Generation in Waveform Space
arXiv
WavFlow: Audio Generation in Waveform Space new
arXiv · 2026-05-18
Modern audio generation predominantly relies on latent-space compression, introducing additional complexity and potential information loss. In this work, we challenge this paradigm with WavFlow, a framework that generates high-fidelity audio directly in raw waveform space without intermediate representations. To overcome the inherent difficulties of modeling high-dimensional and low-energy signals, we reshape audio into 2D token grids through waveform patchify and introduce amplitude lifting to align signal scales, enabling stable optimization via direct x-prediction in flow matching. To capture complex semantic alignment and temporal synchronization, we leverage an automated data pipeline to curate 5 million high-quality video-text-audio triplets, allowing the model to learn fine-grained acoustic patterns from scratch. Experimental results show that WavFlow achieves competitive performa
Research (arXiv)
Aurora: Unified Video Editing with a Tool-Using
arXiv
Aurora: Unified Video Editing with a Tool-Using Agent new
arXiv · 2026-05-18
Recent video editing models have converged on a unified conditioning design: a single diffusion transformer jointly consumes text, source video, and reference images, and one set of weights covers replacement, removal, style transfer, and reference-driven insertion. The design is flexible, but it assumes that the user already provides model-ready text, reference images, and spatial grounding for local edits, which real requests often omit. We present Aurora, an agentic video editing framework that pairs a tool-augmented vision-language model (VLM) agent with a unified video diffusion transformer. The VLM agent maps a raw user request to a structured edit plan aligned with the transformer's conditioning channels, thereby resolving textual and visual underspecification before generation. We train the VLM agent with supervised data for complete edit planning and reference-image selection, t
Research (arXiv)
Code as Agent Harness
arXiv
Code as Agent Harness new
arXiv · 2026-05-18
Recent large language models (LLMs) have demonstrated strong capabilities in understanding and generating code, from competitive programming to repository-level software engineering. In emerging agentic systems, code is no longer only a target output. It increasingly serves as an operational substrate for agent reasoning, acting, environment modeling, and execution-based verification. We frame this shift through the lens of agent harnesses and introduce code as agent harness: a unified view that centers code as the basis for agent infrastructure. To systematically study this perspective, we organize the survey around three connected layers. First, we study the harness interface, where code connects agents to reasoning, action, and environment modeling. Second, we examine harness mechanisms: planning, memory, and tool use for long-horizon execution, together with feedback-driven control a
Research (arXiv)
ESI-Bench: Towards Embodied Spatial Intelligence that Closes
arXiv
ESI-Bench: Towards Embodied Spatial Intelligence that Closes the Perception-Action Loop new
arXiv · 2026-05-18
Spatial intelligence unfolds through a perception-action loop: agents act to acquire observations, and reason about how observations vary as a function of action. Rather than passively processing what is seen, they actively uncover what is unseen - occluded structure, dynamics, containment, and functionality that cannot be resolved from passive sensing alone. We move beyond prior formulations of spatial intelligence that assume oracle observations by recasting the observer as an actor. We introduce ESI-BENCH, a comprehensive benchmark for embodied spatial intelligence spanning 10 task categories and 29 subcategories built on OmniGibson, grounded in Spelke's core knowledge systems. Agents must decide what abilities to deploy - perception, locomotion, and manipulation - and how to sequence them to actively accumulate task-relevant evidence. We conduct extensive experiments on state-of-the-
Research (arXiv)
SURGE: Approximation-free Training Free Particle Filter for
arXiv
SURGE: Approximation-free Training Free Particle Filter for Diffusion Surrogate new
arXiv · 2026-05-18
Diffusion-based generative models increasingly rely on inference-time guidance, adding a drift term or reweighting mixture of experts, to improve sample quality on task-specific objectives. However, most existing techniques require repeated score or gradient evaluations, introducing bias, high computational overhead, or both. We introduce \texttt{URGE}, Unbiased Resampling via Girsanov Estimation, a derivative-free inference-time scaling algorithm that performs path-wise importance reweighting via a Girsanov change of measure. Instead of computing gradient-based particle weights in previous work, \texttt{URGE} attaches a simple multiplicative weight to each simulated trajectory and periodically resamples. No score, no Hessian, and no PDE evaluation is required. We establish an equivalence between path-wise and particle-wise SMC: the Girsanov path weight admits a backward conditional expe
Research (arXiv)
Actionable World Representation
arXiv
Actionable World Representation new
arXiv · 2026-05-18
Inspired by the emergent behaviors in large language models that generalized human intelligence, the research community is pursuing similar emergent capabilities within world models, with a emphasis on modeling the physical world. Within the scope of physical world model, objects are the fundamental primitives that constitute physical reality. From humans to computers, nearly everything we interact with is an object. These objects are rarely static; they are actionable entities with varying states determined by their intrinsic properties. While current methods approach object action states either via video generation or dynamic scene reconstruction, none explicitly model this basic element in a unified, principled way to build an actionable object representation. We propose WorldString, a neural architecture capable of modeling the state manifold of real-world objects by learning directl
Research (arXiv)
Vision-OPD: Learning to See Fine Details for
arXiv
Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation new
arXiv · 2026-05-18
Multimodal Large Language Models (MLLMs) still struggle with fine-grained visual understanding, where answers often depend on small but decisive evidence in the full image. We observe a regional-to-global perception gap: the same MLLM answers fine-grained questions more accurately when conditioned on evidence-centered crops than on the corresponding full images, suggesting that many failures stem from difficulty to focus on relevant evidence rather than insufficient local recognition ability. Motivated by this observation, we propose Vision-OPD (Vision On-Policy Distillation), a regional-to-global self-distillation framework that transfers the model's own privileged regional perception to its full-image policy. Vision-OPD instantiates two conditional policies from the same MLLM: a crop-conditioned teacher and a full-image-conditioned student. The student generates on-policy rollouts, and
Security (NVD + CISA KEV)
14 item(s)
Security (NVD + CISA KEV)
CVE-2017-3324· CVSS 10.0 CRITICAL
NVD
CVE-2017-3324· CVSS 10.0 CRITICAL
NVD · 2017-01-28
Vulnerability in the Primavera P6 Enterprise Project Portfolio Management component of Oracle Primavera Products Suite (subcomponent: Web Access). Supported versions that are affected are 8.2, 8.3, 8.4, 15.1, 15.2, 16.1 and 16.2. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Primavera P6 Enterprise Project Portfolio Management. While the vulnerability is in Primavera P6 Enterprise Project Portfolio Management, attacks may significantly impact additional products. Successful attacks of this vulnerability can result in unauthorized creation, deletion or modification access to critical data or all Primavera P6 Enterprise Project Portfolio Management accessible data as well as unauthorized access to critical data or complete access to all Primavera P6 Enterprise Project Portfolio Management accessible data and unauthorized ability
Security (NVD + CISA KEV)
CVE-2016-3152· CVSS 9.8 CRITICAL
NVD
CVE-2016-3152· CVSS 9.8 CRITICAL
NVD · 2017-01-13
Barco ClickShare CSC-1 devices with firmware before 01.09.03 allow remote attackers to obtain the root password by downloading and extracting the firmware image.
Security (NVD + CISA KEV)
CVE-2016-9299· CVSS 9.8 CRITICAL
NVD
CVE-2016-9299· CVSS 9.8 CRITICAL
NVD · 2017-01-13
The remoting module in Jenkins before 2.32 and LTS before 2.19.3 allows remote attackers to execute arbitrary code via a crafted serialized Java object, which triggers an LDAP query to a third-party server.
Security (NVD + CISA KEV)
CVE-2016-10141· CVSS 9.8 CRITICAL
NVD
CVE-2016-10141· CVSS 9.8 CRITICAL
NVD · 2017-01-13
An integer overflow vulnerability was observed in the regemit function in regexp.c in Artifex Software, Inc. MuJS before fa3d30fd18c348bb4b1f3858fb860f4fcd4b2045. The attack requires a regular expression with nested repetition. A successful exploitation of this issue can lead to code execution or a denial of service (buffer overflow) condition.
Security (NVD + CISA KEV)
CVE-2015-3188· CVSS 9.8 CRITICAL
NVD
CVE-2015-3188· CVSS 9.8 CRITICAL
NVD · 2017-01-13
The UI daemon in Apache Storm 0.10.0 before 0.10.0-beta1 allows remote attackers to execute arbitrary code via unspecified vectors.
Security (NVD + CISA KEV)
CVE-2016-2090· CVSS 9.8 CRITICAL
NVD
CVE-2016-2090· CVSS 9.8 CRITICAL
NVD · 2017-01-13
Off-by-one vulnerability in the fgetwln function in libbsd before 0.8.2 allows attackers to have unspecified impact via unknown vectors, which trigger a heap-based buffer overflow.
Security (NVD + CISA KEV)
CVE-2016-8204· CVSS 9.8 CRITICAL
NVD
CVE-2016-8204· CVSS 9.8 CRITICAL
NVD · 2017-01-15
A Directory Traversal vulnerability in FileReceiveServlet in the Brocade Network Advisor versions released prior to and including 14.0.2 could allow remote attackers to upload a malicious file in a section of the file system where it can be executed.
Security (NVD + CISA KEV)
CVE-2016-8205· CVSS 9.8 CRITICAL
NVD
CVE-2016-8205· CVSS 9.8 CRITICAL
NVD · 2017-01-15
A Directory Traversal vulnerability in DashboardFileReceiveServlet in the Brocade Network Advisor versions released prior to and including 14.0.2 could allow remote attackers to upload a malicious file in a section of the file system where it can be executed.
Security (NVD + CISA KEV)
CVE-2017-5517· CVSS 9.8 CRITICAL
NVD
CVE-2017-5517· CVSS 9.8 CRITICAL
NVD · 2017-01-17
SQL injection vulnerability in author.control.php in GeniXCMS through 0.0.8 allows remote attackers to execute arbitrary SQL commands via the type parameter.
Security (NVD + CISA KEV)
CVE-2017-5519· CVSS 9.8 CRITICAL
NVD
CVE-2017-5519· CVSS 9.8 CRITICAL
NVD · 2017-01-17
SQL injection vulnerability in Posts.class.php in GeniXCMS through 0.0.8 allows remote attackers to execute arbitrary SQL commands via the id parameter.
Security (NVD + CISA KEV)
CVE-2016-7996· CVSS 9.8 CRITICAL
NVD
CVE-2016-7996· CVSS 9.8 CRITICAL
NVD · 2017-01-18
Heap-based buffer overflow in the WPG format reader in GraphicsMagick 1.3.25 and earlier allows remote attackers to have unspecified impact via a colormap with a large number of entries.
Security (NVD + CISA KEV)
CVE-2016-9676· CVSS 9.8 CRITICAL
NVD
CVE-2016-9676· CVSS 9.8 CRITICAL
NVD · 2017-01-19
Buffer overflow in Citrix Provisioning Services before 7.12 allows attackers to execute arbitrary code via unspecified vectors.
Security (NVD + CISA KEV)
CVE-2016-9678· CVSS 9.8 CRITICAL
NVD
CVE-2016-9678· CVSS 9.8 CRITICAL
NVD · 2017-01-19
Use-after-free vulnerability in Citrix Provisioning Services before 7.12 allows attackers to execute arbitrary code via unspecified vectors.
Security (NVD + CISA KEV)
CVE-2016-9679· CVSS 9.8 CRITICAL
NVD
CVE-2016-9679· CVSS 9.8 CRITICAL
NVD · 2017-01-19
Citrix Provisioning Services before 7.12 allows attackers to execute arbitrary code by overwriting a function pointer.
Projects + Resources (Discovery)
10 item(s)
Projects + Resources (Discovery)
ROS2 Tutorials - YouTube
Brave Search
ROS2 Tutorials - YouTube new
Brave Search
Learn ROS2, 5 minutes at a time
Projects + Resources (Discovery)
ROS2 Tutorial - ROS2 Humble 2H50 [Crash
Brave Search
ROS2 Tutorial - ROS2 Humble 2H50 [Crash Course] - YouTube new
Brave Search
This ROS2 Tutorial [Crash Course] will get you started with ROS2 Humble in no time. Step by step and hands-on lessons only! You will learn how to work with R...
Projects + Resources (Discovery)
JavaScript 教程 - 菜鸟教程
Brave Search
JavaScript 教程 - 菜鸟教程 new
Brave Search
什么是JavaScript JavaScript是动态实现Web页面效果的脚本语言，是Web前端必须掌握的三种语言之一。 用于定义网页内容的HTML CSS指定网页的布局 用于编程网页行为的JavaScript Web网页不是唯一使用JavaScript的地方。许多桌面和服务器程序都使用Java ...
Projects + Resources (Discovery)
JavaScript 标准参考教程（alpha） -- JavaScript 标准参考教程（alpha）
Brave Search
JavaScript 标准参考教程（alpha） -- JavaScript 标准参考教程（alpha） new
Brave Search
英文标题：JavaScript Standards Reference Guide · 授权方式：创意共享“署名-非商业性使用”许可证
Projects + Resources (Discovery)
Jiacheng Lin’s Homepage
Brave Search
Jiacheng Lin’s Homepage new
Brave Search
My research focuses on foundation models (e.g., large language models (LLMs), multi-modal models), reinforcement learning (RL), and agents, with their applications in healthcare, biomedicine, and recommendation systems.
Projects + Resources (Discovery)
Jiacheng (James) Zhang - Homepage
Brave Search
Jiacheng (James) Zhang - Homepage new
Brave Search
Hi there, I am Jiacheng (James) Zhang. I am an associate member of Sea AI Lab (SAIL), mentored by Dr. Tianyu Pang, and a Ph.D. candidate at Trustworthy Machine Learning and Reasoning (TMLR) group in the Faculty of Engineering and Information Technology, the University of Melbourne, advised by Dr.
Projects + Resources (Discovery)
piwheels - w3school-python-jiaocheng-fix1
Brave Search
piwheels - w3school-python-jiaocheng-fix1 new
Brave Search
The piwheels project page for w3school-python-jiaocheng-fix1: W3School Python 教程
Projects + Resources (Discovery)
Client Challenge
Brave Search
Client Challenge new
Brave Search
JavaScript is disabled in your browser · Please enable JavaScript to proceed · A required part of this site couldn’t load. This may be due to a browser extension, network issues, or browser settings. Please check your connection, disable any ad blockers, or try using a different browser
Projects + Resources (Discovery)
Jiahao Chen new
Brave Search
I received my bachelor's and master's degrees from Harbin Institute of Technology. <strong>My research focuses on reliable and adaptable machine learning under realistic settings, with particular interests in long-tailed learning, representation learning, and AI4Science</strong>.
Projects + Resources (Discovery)
Learn JavaScript – Free JS Courses for
Brave Search
Learn JavaScript – Free JS Courses for Beginners new
Brave Search
JavaScript is a popular programming language used for building web applications, among many other things. In this article, I will provide a list of free online resources where you can start learning JavaScript. After you have learned the basics then...
AI Power Progress iA