Tech + AI + Science NewsTech + AI + Science News
14 item(s)
Tech + AI + Science News
The next phase of OpenAI’s Education for
OpenAI News
The next phase of OpenAI’s Education for Countries
OpenAI News · 2026-05-20
OpenAI advances Education for Countries, expanding AI adoption in schools with new partnerships, teacher training, and tools to improve global learning outcomes.
Tech + AI + Science News
How Ramp engineers accelerate code review with
OpenAI News
How Ramp engineers accelerate code review with Codex
OpenAI News · 2026-05-20
How Ramp engineers use Codex with GPT-5.5 to review code and ship improvements, allowing them to get substantive feedback in minutes instead of hours.
Tech + AI + Science News
Introducing OpenAI for Singapore
OpenAI News
Introducing OpenAI for Singapore
OpenAI News · 2026-05-19
OpenAI for Singapore launches a multi-year AI partnership to expand deployment, build local talent, and support businesses and public services with AI.
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
5 open-source developer tools that are better
Brave News
5 open-source developer tools that are better than their well-funded competitors new
Brave News
As someone who frequently maps ... brainstorming a new project layout on an infinite canvas (like dragging in blocks, drawing connectors, and grouping ideas), and then flip to a document view to refine the technical specs. AFFiNE is an open-source PKM tool that rivals ...
Tech + AI + Science News
New $11M Effort Aims to Train Teachers
Brave News
New $11M Effort Aims to Train Teachers in AI. How Does It Work? new
Brave News
As part of the grant, CSTA will ... learning and ongoing support. “There’s not very much research that [explores] how teachers actually do [integrate AI concepts] in meaningful and ongoing ways, so we want to study that,” Twarek said. ... Lauraine Langreo is an Education Week staff writer, covering education technology and learning ...
Tech + AI + Science News
Drupal to Patch Highly Critical Vulnerability at
Brave News
Drupal to Patch Highly Critical Vulnerability at Risk of Quick Exploitation - SecurityWeek
Brave News
Drupal is warning users that it’s preparing a patch for a ‘highly critical’ vulnerability that may be exploited shortly after its disclosure.
Tech + AI + Science News
NKGen Biotech Expands Clinical Trial Site Network
Brave News
NKGen Biotech Expands Clinical Trial Site Network for Alzheimer’s Disease Program Across U.S. to Improve Patient Access and Support Enrollment new
Brave News
New sites in New Jersey, New York, and Florida broaden geographic reach and increase access for patients participating in Phase 2 trial of troculeucel in...
Tech + AI + Science News
Python 3.15: features that didn't make the
Hacker News
Python 3.15: features that didn't make the headlines new
Hacker News · 2026-05-21
Comments
Tech + AI + Science News
Google officially announces that ads will be
Hacker News
Google officially announces that ads will be included in AI Mode search results new
Hacker News · 2026-05-21
Comments
Tech + AI + Science News
No Slop Grenade
Hacker News
No Slop Grenade new
Hacker News · 2026-05-21
Comments
Tech + AI + Science News
Show HN: Rmux – A programmable terminal
Hacker News
Show HN: Rmux – A programmable terminal multiplexer with a Playwright-style SDK new
Hacker News · 2026-05-21
Comments
Tech + AI + Science News
Investigating unauthorized access to GitHub-owned repositories
GitHub Blog
Investigating unauthorized access to GitHub-owned repositories
GitHub Blog · 2026-05-20
If any impact is discovered, customers will be notified via established incident response and notification channels. The post Investigating unauthorized access to GitHub-owned repositories appeared first on The GitHub Blog .
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 .
Research (arXiv)
10 item(s)
Research (arXiv)
Variance Reduction for Expectations with Diffusion Teachers
arXiv
Variance Reduction for Expectations with Diffusion Teachers new
arXiv · 2026-05-20
Pretrained diffusion models serve as frozen teachers feeding downstream pipelines such as text-to-3D, single-step distillation, and data attribution. The teacher gradients these pipelines consume are Monte Carlo (MC) expectations over noise levels and Gaussian noise samples; their estimator variance dominates compute cost because each draw requires expensive upstream work (rendering, simulation, encoding). We introduce CARV, a compute-aware variance-accounting framework that motivates a hierarchical MC estimator: amortize the expensive upstream computation over cheap diffusion-noise resamples, sharpened by timestep importance sampling and a stratified-inverse-CDF construction. In our text-to-3D distillation and attribution experiments, CARV delivers 2-3x effective compute multipliers (most from amortized reuse; ~25% additional from IS+stratification) without changing the objective; in si
Research (arXiv)
Equilibrium Reasoners: Learning Attractors Enables Scalable Rea…
arXiv
Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning new
arXiv · 2026-05-20
Scaling test-time compute by iteratively updating a latent state has emerged as a powerful paradigm for reasoning. Yet the internal mechanisms that enable these iterative models to generalize beyond memorized patterns remain unclear. We hypothesize that generalizable reasoning arises from learning task-conditioned attractors: latent dynamical systems whose stable fixed points correspond to valid solutions. We formalize this process through Equilibrium Reasoners (EqR), which enable test-time scaling without external verifiers or task-specific priors. EqR scales internal dynamics along two axes: depth, by running more iterations, and breadth, by aggregating stochastic trajectories from multiple initializations. Empirically, gains from test-time scaling are tightly coupled with stronger convergence toward solution-aligned attractors. This attractor perspective allows neural networks to adap
Research (arXiv)
Uni-Edit: Intelligent Editing Is A General Task
arXiv
Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning new
arXiv · 2026-05-20
Currently, enhancing Unified Multimodal Models (UMMs) with image understanding, generation, and editing capabilities mainly relies on mixed multi-task training. Due to inherent task conflicts, such strategy requires complex multi-stage pipelines, massive data mixing, and balancing tricks, merely resulting in a performance trade-off rather than true mutual reinforcement. To break this paradigm, we propose Uni-Edit, an intelligent image editing task that serves as the first general task for UMM tuning. Unlike complex mixed pipelines, Uni-Edit improves performance across all three abilities at once using only one task, one training stage, and one dataset. Specifically, we first identify image editing as an inherently ideal general task, as it naturally demands both visual understanding and generation. However, existing editing data relies on simplistic instructions that severely underutiliz
Research (arXiv)
Quantifying Hyperparameter Transfer and the Importance of
arXiv
Quantifying Hyperparameter Transfer and the Importance of Embedding Layer Learning Rate new
arXiv · 2026-05-20
Hyperparameter transfer allows extrapolating optimal optimization hyperparameters from small to large scales, making it critical for training large language models (LLMs). This is done either by fitting a scaling law to the hyperparameters or by a judicious choice of parameterization, such as Maximal Update ($μ$P), that renders optimal hyperparameters approximately scale invariant. In this paper, we first develop a framework to quantify hyperparameter transfer through three metrics: (1) the quality of the scaling law fit, (2) the robustness to extrapolation errors, and (3) the asymptotic loss penalty due to choice of parameterization. Next, we investigate through a comprehensive series of ablations why $μ$P appears to offer high-quality learning rate transfer relative to standard parameterization (SP), as existing theory is inadequate. We find that the overwhelming benefit of $μ$P relati
Research (arXiv)
EvoStruct: Bridging Evolutionary and Structural Priors for
arXiv
EvoStruct: Bridging Evolutionary and Structural Priors for Antibody CDR Design via Protein Language Model Adaptation new
arXiv · 2026-05-20
Equivariant graph neural network (GNN) methods for antibody complementarity-determining region (CDR) design achieve the highest sequence recovery but suffer from severe vocabulary collapse. The current best GNN methods over-predict very few amino acids, such as tyrosine and glycine, while ignoring functionally important residues. We trace this failure to GNN encoders learning amino acid distributions de novo from limited structural data, discarding substitution patterns encoded in evolutionary databases. To resolve this, we propose EvoStruct, which bridges a frozen protein language model (PLM) with 3D structural context from an E(3)-equivariant GNN via a cross-attention adapter. Unlike prior PLM-structure adapters for general protein design, EvoStruct targets the vocabulary collapse problem specific to CDR design through progressive PLM unfreezing and R-Drop consistency regularization. O
Research (arXiv)
One-Step Distillation of Discrete Diffusion Image Generators
arXiv
One-Step Distillation of Discrete Diffusion Image Generators via Fixed-Point Iteration new
arXiv · 2026-05-20
Discrete diffusion models excel at visual synthesis but rely on slow, iterative decoding. Existing single-step distillation methods attempt to bypass this bottleneck, either by training auxiliary score networks that effectively double compute, or by introducing specialized parameterizations and multi-stage pipelines that fragment optimization. In this paper, we introduce Fixed-Point Distillation (FPD), an end-to-end framework that constructs local correction targets by partially corrupting the student's one-step draft and refining it with a single teacher step. To compute the training objective in a semantically meaningful space, we lift discrete tokens into continuous features and apply a multi-bandwidth drift loss that iteratively accumulates these corrections. To backpropagate through the discrete bottleneck, we employ a straight-through estimator that feeds exact hard-sampled tokens
Research (arXiv)
Velocityformer: Broken-Symmetry-Matched Equivariant Graph Trans…
arXiv
Velocityformer: Broken-Symmetry-Matched Equivariant Graph Transformers for Cosmological Velocity Reconstruction new
arXiv · 2026-05-20
Precise measurement of the kinematic Sunyaev-Zel'dovich (kSZ) effect - a probe of the large-scale distribution of baryonic matter, a key observable for cosmological inference - requires accurate reconstruction of galaxy velocities from spectroscopic surveys. The signal-to-noise ratio (SNR) of kSZ measurements scales directly with the correlation coefficient $r$ between reconstructed and true velocities. We introduce Velocityformer, an equivariant graph transformer architecture designed to match the specific symmetry of the observational data. While the underlying physics is equivariant with respect to translations and rotations, observational effects break this symmetry due to the preferred line-of-sight direction. Matching the model's inductive bias to the data's broken symmetry consistently improves performance across all model sizes and training volumes, with Velocityformer improving
Research (arXiv)
DeepWeb-Bench: A Deep Research Benchmark Demanding Massive
arXiv
DeepWeb-Bench: A Deep Research Benchmark Demanding Massive Cross-Source Evidence and Long-Horizon Derivation new
arXiv · 2026-05-20
Deep research, in which an agent searches the open web, collects evidence, and derives an answer through extended reasoning, is a prominent use case for frontier language models. Frontier deep research products score high on existing benchmarks, making it difficult to distinguish their capabilities from current evaluation data alone. We introduce DeepWeb-Bench, a deep research benchmark that is substantially harder than existing benchmarks for the current frontier. Difficulty comes from three properties of the data itself: each task requires massive evidence collection, cross-source reconciliation, and long-horizon multi-step derivation. We represent these three sources of difficulty as four capability families (Retrieval, Derivation, Reasoning, and Calibration) and report results sliced by family. Every reference answer is accompanied by a source-provenance record with four disclosure l
Research (arXiv)
AiraXiv: An AI-Driven Open-Access Platform for Human
arXiv
AiraXiv: An AI-Driven Open-Access Platform for Human and AI Scientists new
arXiv · 2026-05-20
Recent advances in artificial intelligence (AI) have accelerated the growth of both human-authored and AI-generated research outputs, placing increasing strain on traditional academic publishing systems and challenging the scalability of conference- and journal-centered paradigms amid rising submission volumes, reviewer workload, and venue size. To address these challenges, we explore an AI-era publishing paradigm in which both human and AI scientists participate as authors and readers, and papers evolve through continuous, feedback-driven iteration. We propose AiraXiv, an AI-driven open-access platform built on open preprints, AI-augmented analysis and review, and reader feedback. AiraXiv supports human scientists through an interactive UI and AI scientists through Model Context Protocol (MCP)-based interactions. We validate AiraXiv through real-world deployments, including serving as t
Research (arXiv)
WikiVQABench: A Knowledge-Grounded Visual Question Answering Be…
arXiv
WikiVQABench: A Knowledge-Grounded Visual Question Answering Benchmark from Wikipedia and Wikidata new
arXiv · 2026-05-20
Visual Question Answering (VQA) benchmarks have largely emphasized perception-based tasks that can be solved from visual content alone. In contrast, many real-world scenarios require external knowledge that is not directly observable in the image to answer correctly. We introduce WikiVQABench, a human-curated knowledge-grounded VQA benchmark constructed by systematically combining Wikipedia images, their associated article captions, and structured knowledge from Wikidata. Our pipeline uses large language models (LLMs) to generate candidate multiple-choice image-question-answer sets. All generated instances are subsequently reviewed and curated by human annotators to ensure factual correctness, visual-text consistency, and that each question requires external knowledge in addition to visual evidence for correct resolution. WikiVQABench comprises a substantial collection of Wikipedia image
Security (NVD + CISA KEV)
14 item(s)
Security (NVD + CISA KEV)
CVE-2026-37541· CVSS 10.0 CRITICAL
NVD
CVE-2026-37541· CVSS 10.0 CRITICAL
NVD · 2026-05-01
Buffer overflow vulnerability in Open Vehicle Monitoring System 3 (OVMS3) 3.3.005. In canformat_gvret.cpp, the length field in GVRET binary data is not properly validated, allowing remote attackers to cause a denial of service or possibly execute arbitrary code via crafted GVRET frames.
Security (NVD + CISA KEV)
CVE-2023-4670· CVSS 9.8 CRITICAL
NVD
CVE-2023-4670· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Innosa Probbys allows SQL Injection. This issue affects Probbys: before 2.
Security (NVD + CISA KEV)
CVE-2023-5046· CVSS 9.8 CRITICAL
NVD
CVE-2023-5046· CVSS 9.8 CRITICAL new
NVD · 2023-10-12
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Biltay Technology Procost allows SQL Injection, Command Line Execution through SQL Injection. This issue affects Procost: before 1390.
Security (NVD + CISA KEV)
CVE-2023-5047· CVSS 9.8 CRITICAL
NVD
CVE-2023-5047· CVSS 9.8 CRITICAL new
NVD · 2023-11-22
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in DRD Fleet Leasing DRDrive allows SQL Injection. This issue affects DRDrive: before 20231006.
Security (NVD + CISA KEV)
CVE-2023-4832· CVSS 9.8 CRITICAL
NVD
CVE-2023-4832· CVSS 9.8 CRITICAL new
NVD · 2023-09-14
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Aceka Company Management allows SQL Injection. This issue affects Company Management: before 3072 .
Security (NVD + CISA KEV)
CVE-2023-4766· CVSS 9.8 CRITICAL
NVD
CVE-2023-4766· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Movus allows SQL Injection. This issue affects Movus: before 20230913.
Security (NVD + CISA KEV)
CVE-2023-4972· CVSS 9.8 CRITICAL
NVD
CVE-2023-4972· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Incorrect Use of Privileged APIs vulnerability in Yepas Digital Yepas allows Collect Data as Provided by Users. This issue affects Digital Yepas: before 1.0.1.
Security (NVD + CISA KEV)
CVE-2023-4830· CVSS 9.8 CRITICAL
NVD
CVE-2023-4830· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Tura Signalix allows SQL Injection. This issue affects Signalix: 7T_0228.
Security (NVD + CISA KEV)
CVE-2023-4833· CVSS 9.8 CRITICAL
NVD
CVE-2023-4833· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Besttem Network Marketing Software allows SQL Injection. This issue affects Network Marketing Software: before 1.0.2309.6.
Security (NVD + CISA KEV)
CVE-2023-4835· CVSS 9.8 CRITICAL
NVD
CVE-2023-4835· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in CF Software Oil Management Software allows SQL Injection. This issue affects Oil Management Software: before 20230912 .
Security (NVD + CISA KEV)
CVE-2023-5045· CVSS 9.8 CRITICAL
NVD
CVE-2023-5045· CVSS 9.8 CRITICAL new
NVD · 2023-10-12
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Biltay Technology Kayisi allows SQL Injection, Command Line Execution through SQL Injection. This issue affects Kayisi: before 1286.
Security (NVD + CISA KEV)
CVE-2023-4702· CVSS 9.8 CRITICAL
NVD
CVE-2023-4702· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Authentication Bypass Using an Alternate Path or Channel vulnerability in Yepas Digital Yepas allows Authentication Bypass. This issue affects Digital Yepas: before 1.0.1.
Security (NVD + CISA KEV)
CVE-2023-4673· CVSS 9.8 CRITICAL
NVD
CVE-2023-4673· CVSS 9.8 CRITICAL new
NVD · 2023-09-15
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Sanalogy Turasistan allows SQL Injection. This issue affects Turasistan: before 20230911 .
Security (NVD + CISA KEV)
CVE-2023-4737· CVSS 9.8 CRITICAL
NVD
CVE-2023-4737· CVSS 9.8 CRITICAL new
NVD · 2023-09-27
Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability in Hedef Tracking Admin Panel allows SQL Injection. This issue affects Admin Panel: before 1.2.
Projects + Resources (Discovery)
10 item(s)
Projects + Resources (Discovery)
Curso de Python [Gratis y Certificado]
Brave Search
Curso de Python [Gratis y Certificado] new
Brave Search
He podido practicar y recordar lo que sabía de Python, es muy claro el profesor en explicar, excelente lo recomiendo, muchas gracias por las clases. ... Claro que si, todos los cursos disponibles en <strong>Edutin Academy</strong> son de acceso gratis.
Projects + Resources (Discovery)
Tous les MOOC dédiés au Machine Learning
Brave Search
Tous les MOOC dédiés au Machine Learning ou Apprentissage Automatique | My Mooc new
Brave Search
Tu trouveras toutes les réponses à tes questions en suivant les formations gratuites proposées ici. Avec ces cours en ligne en apprentissage automatique, tu peux suivre une formation sur le langage Python, sur le langage R, ou même sur le concept de réseau de neurones artificiels.
Projects + Resources (Discovery)
Meilleurs cours et tutoriels gratuits sur le
Brave Search
Meilleurs cours et tutoriels gratuits sur le sujet Machine Learning en ligne - Mise à jour : [février 2026] new
Brave Search
Apprenez l'apprentissage automatique des formateurs les mieux notés. Trouvez les meilleurs cours d'apprentissage automatique pour votre niveau et vos besoins, allant de l'analyse du Big Data et de la modélisation des données aux algorithmes d'apprentissage automatique, aux réseaux neuronaux, ...
Projects + Resources (Discovery)
Meilleurs cours et certificats en Apprentissage automatique
Brave Search
Meilleurs cours et certificats en Apprentissage automatique 2026 | Coursera new
Brave Search
Les cours en apprentissage automatique peuvent vous aider à découvrir modèles, données et techniques d'évaluation. Comparez les options. Inscription gratuite.
Projects + Resources (Discovery)
Apprenez les bases du langage Python -
Brave Search
Apprenez les bases du langage Python - OpenClassrooms new
Brave Search
Dans ce cours, vous allez enregistrer des informations en tant que données en Python. Vous allez contrôler la logique dans votre code avec des conditions, des boucles et des fonctions. Vous allez aussi automatiser la récupération de données à partir du web avec les packages Requests, Beautiful Soup et CSV, dans le processus appelé extraction de données web.
Projects + Resources (Discovery)
Learn Python - Free Interactive Python Tutorial
Brave Search
Learn Python - Free Interactive Python Tutorial new
Brave Search
learnpython.org is a free interactive Python tutorial for people who want to learn Python, fast.
Projects + Resources (Discovery)
Meilleurs cours et certificats en Python 2026
Brave Search
Meilleurs cours et certificats en Python 2026 new
Brave Search
Please use a modern browser with JavaScript enabled to use Coursera.
Projects + Resources (Discovery)
Apprenez à programmer avec JavaScript - OpenClassrooms
Brave Search
Apprenez à programmer avec JavaScript - OpenClassrooms new
Brave Search
Maîtrisez les bases de la programmation JavaScript et sa logique de programmation en réalisant un site web dynamique. ... Les applications s'appuient sur des données et une logique pour accomplir leur magie. Cela exige qu'un programmeur – vous – indique à l'ordinateur ce qu'il faut faire et comment. Alors, comment écrire une application ? Par où commencer ? Dans ce cours, vous découvrirez les bases du langage de programmation JavaScript : sa syntaxe, sa logique de programmation…
Projects + Resources (Discovery)
Meilleurs cours et tutoriels gratuits sur le
Brave Search
Meilleurs cours et tutoriels gratuits sur le sujet JavaScript en ligne - Mise à jour : [février 2026] new
Brave Search
Apprenez Javascript avec les formateurs les mieux notés. Trouvez les meilleures cours de Javascript en ligne et commencez à coder en Javascript dès aujourd'hui. Javascript est un langage de programmation informatique orienté objet couramment utilisé par les développeurs Web et les ...
Projects + Resources (Discovery)
Le Tutoriel JavaScript Moderne
Brave Search
Le Tutoriel JavaScript Moderne new
Brave Search
Tutoriel JavaScript moderne : explications simples mais détaillées avec des exemples et des exercices, y compris : les closures, le document et les événements, la programmation orientée objet et plus encore.
AI Power Progress iA