Skip to content
Provenance Brief
Provenance Brief

Browse

59 stories from 58 sources

Press just now

In an interview with WIRED, Greg Brockman says supporting…OpenAI’s President Gave Millions to Trump. He Says It’s for Humanity

In an interview with WIRED, Greg Brockman says supporting politicians who back AI is…

Verify with primary source before acting.

Wired AI
Research just now

OpenAI is shutting down GPT-4o and several other older models in…OpenAI is retiring GPT-4o and three other legacy models tomorrow, likely for good

OpenAI is shutting down GPT-4o and several other older models in ChatGPT tomrrow.

Find the core claim, method, and released artifacts. Also: note model size and…

The Decoder
Tech just now

On October 24th, 2011, Jeffrey Epstein was emailing back and forth…Jeffrey Epstein might not have created /pol/, but he helped carry out its mission

On October 24th, 2011, Jeffrey Epstein was emailing back and forth with Boris Nikolic,…

Verify with primary source before acting. Also: check API docs for breaking changes.

The Verge Tech
Business just now

OpenAI on Thursday launched GPT-5.3-Codex-Spark , a stripped-down…OpenAI deploys Cerebras chips for 15x faster code generation in first major move beyond Nvidia

OpenAI on Thursday launched GPT-5.3-Codex-Spark , a stripped-down coding model…

Track implications for your stack/vendors. Also: check API docs for breaking changes;…

VentureBeat
Product just now

Interrupt - The Agent Conference by LangChain - is where builders…Join us for Interrupt: The Agent Conference

Interrupt - The Agent Conference by LangChain - is where builders come to learn what's…

Check docs/changelog for breaking changes. Also: check API docs for breaking changes.

LangChain Blog
Research just now

Google Deepmind says its upgraded Gemini 3 Deep Think mode leads…Google Deepmind upgrades Gemini 3 Deep Think for complex science and engineering tasks

Google Deepmind says its upgraded Gemini 3 Deep Think mode leads major reasoning and…

Find the core claim, method, and released artifacts. Also: verify benchmark methodology.

The Decoder
Research 2h ago

Google Deepmind's AI agent Aletheia independently wrote a math…Deepmind's research AI occasionally solves what humans can't and mostly gets everything else wrong

Google Deepmind's AI agent Aletheia independently wrote a math paper, disproved a…

Find the core claim, method, and released artifacts. Also: verify benchmark methodology.

The Decoder
Tech 2h ago

The hours-long closure of El Paso airspace stemmed from the use of…El Paso airspace closure was reportedly triggered by the CBP’s use of an anti-drone laser

The hours-long closure of El Paso airspace stemmed from the use of an anti-drone laser…

Verify with primary source before acting. Also: check LICENSE and dependencies.

The Verge Tech
Tech 2h ago

In a style that feels nostalgic and nascent, Bree O'Donnell is…Every little thing she does is magic

In a style that feels nostalgic and nascent, Bree O'Donnell is crafting her singular…

Verify with primary source before acting.

The Verge Tech
Tech 2h ago

The Apple Vision Pro is finally getting an official visionOS…YouTube is coming to the Apple Vision Pro

The Apple Vision Pro is finally getting an official visionOS YouTube app on Thursday.

Verify with primary source before acting.

The Verge Tech
Research 2h ago

The Pentagon is pressing leading AI companies including OpenAI,…Pentagon pushes AI companies to deploy unrestricted models on classified military networks

The Pentagon is pressing leading AI companies including OpenAI, Anthropic, Google, and…

Find the core claim, method, and released artifacts. Also: note model size and…

The Decoder
Research 2h ago

Published on February 12, 2026 4:55 PM GMT As AI systems get…How do we (more) safely defer to AIs?

Published on February 12, 2026 4:55 PM GMT As AI systems get more capable, it becomes…

Note scope, timelines, compliance requirements. Also: check API docs for breaking…

AI Alignment Forum
Research 2h ago

The company plans to fully absorb grid upgrade costs, invest in new…Anthropic promises to cover consumer electricity costs from new data center construction

The company plans to fully absorb grid upgrade costs, invest in new power generation,…

Find the core claim, method, and released artifacts. Also: calculate cost at your scale.

The Decoder
Tech 2h ago

This is Lowpass by Janko Roettgers, a newsletter on the…How Gorilla Tag is weathering the VR winter

This is Lowpass by Janko Roettgers, a newsletter on the ever-evolving intersection of…

Verify with primary source before acting.

The Verge Tech
Business 2h ago

When an AI agent visits a website, it’s essentially a tourist who…Google Chrome ships WebMCP in early preview, turning every website into a structured tool for AI agents

When an AI agent visits a website, it’s essentially a tourist who doesn’t speak the…

Track implications for your stack/vendors. Also: check API docs for breaking changes;…

VentureBeat
Tech 2h ago

Today, we’re going to talk about the role AI might play in…The surprising case for AI judges

Today, we’re going to talk about the role AI might play in deciding legal disputes.

Verify with primary source before acting.

The Verge Tech
Labs 3h ago 2 sources

Our most specialized reasoning mode is now updated to solve modern…Gemini 3 Deep Think: Advancing science, research and engineering

Our most specialized reasoning mode is now updated to solve modern science, research…

Find the core claim, method, and released artifacts.

Google DeepMind
Tech 3h ago

Though it's going to be a while before we see more of…Spider-Noir looks like a hard-boiled thriller in first trailer

Though it's going to be a while before we see more of Sony's animated Spider-Verse or…

Verify with primary source before acting.

The Verge Tech
Research 3h ago

This article is part of our exclusive IEEE Journal Watch series in…ChatGPT’s Translation Skills Parallel Most Human Translators

This article is part of our exclusive IEEE Journal Watch series in partnership with…

Find the core claim, method, and released artifacts. Also: verify benchmark…

IEEE Spectrum AI
Tech 3h ago

Apple, Bose, and Sony are in a running battle for the best…The Sony WF-1000XM6 earbuds reclaim the noise-canceling crown

Apple, Bose, and Sony are in a running battle for the best noise-canceling earbuds, and…

Verify with primary source before acting.

The Verge Tech
Business 3h ago

Lowering the cost of inference is typically a combination of…AI inference costs dropped up to 10x on Nvidia's Blackwell — but hardware is only half the equation

Lowering the cost of inference is typically a combination of hardware and software.

Track implications for your stack/vendors. Also: check API docs for breaking changes;…

VentureBeat
News 4h ago

Moving beyond the black box to turn complex model outputs into…How to Leverage Explainable AI for Better Business Decisions

Moving beyond the black box to turn complex model outputs into actionable…

Verify with primary source before acting. Also: note model size and inference…

Towards Data Science
Research 4h ago

Key Takeaways AI literacy involves understanding how to effectively…AI Literacy For Startups and Small Businesses: Applications + Tips

Key Takeaways AI literacy involves understanding how to effectively and ethically apply…

Check docs/changelog for breaking changes. Also: check API docs for breaking changes;…

Salesforce AI Research
Research 4h ago

Most Salesforce environments don’t exist in isolationArchitecting Inbound Governance with External Client Apps

Most Salesforce environments don’t exist in isolation.

Check docs/changelog for breaking changes. Also: check API docs for breaking changes;…

Salesforce AI Research
Business 4h ago

While the move shows the vendor's trust in its technology, it could…ElevenLabs Insures Agents, Targeting Enterprises' Fears

While the move shows the vendor's trust in its technology, it could create a false…

Track implications for your stack/vendors.

AI Business
Tech 5h ago

Amazon's rolling out a new "Send to Alexa Plus" feature to the…Amazon’s Send to Alexa Plus makes the Kindle Scribe feel more like a productivity device

Amazon's rolling out a new "Send to Alexa Plus" feature to the latest Kindle Scribe and…

Verify with primary source before acting.

The Verge Tech
Press 6h ago

This is today’s edition of The Download , our weekday newsletter that…The Download: AI-enhanced cybercrime, and secure AI assistants

This is today’s edition of The Download , our weekday newsletter that provides a daily…

Verify with primary source before acting. Also: check API docs for breaking changes;…

MIT Technology Review
Research 6h ago

Moltbook , the first social network for generative AI agents , went…The First Social Network for AI Agents Heralds Their Messy Future

Moltbook , the first social network for generative AI agents , went live on 28 January,…

Find the core claim, method, and released artifacts. Also: check API docs for breaking…

IEEE Spectrum AI
Press 6h ago

The use of cryptocurrency in sales of human beings for prostitution…Crypto-Funded Human Trafficking Is Exploding

The use of cryptocurrency in sales of human beings for prostitution and scam compounds…

Verify with primary source before acting. Also: watch for API/product continuity.

Wired AI
News 6h ago

Learn how CPU and GPUs interact in the host-device paradigmAI in Multiple GPUs: Understanding the Host and Device Paradigm

Learn how CPU and GPUs interact in the host-device paradigm

Verify with primary source before acting.

Towards Data Science
Product 6h ago

Browserbase is now available on the Vercel Marketplace, allowing…Browserbase joins the Vercel Agent Marketplace

Browserbase is now available on the Vercel Marketplace, allowing teams to run browser…

Check docs/changelog for breaking changes. Also: check API docs for breaking changes;…

Vercel Blog
Product 6h ago

MiniMax M2.5 is now available on AI Gateway.M2.5 plans before it…Use MiniMax M2.5 on AI Gateway

MiniMax M2.5 is now available on AI Gateway.M2.5 plans before it builds, breaking down…

Check docs/changelog for breaking changes. Also: check API docs for breaking changes;…

Vercel Blog
Product 6h ago

Any new deployment containing a version of the third-party package…New deployments with vulnerable versions of the third-party package next-mdx-remote are now blocked by default

Any new deployment containing a version of the third-party package next-mdx-remote that…

Check docs/changelog for breaking changes. Also: audit your exposure immediately.

Vercel Blog
Tech 6h ago

If you have a spare $7,999 (plus a $250 deposit), hate folding…This $7,999 robot will fold (some of) your laundry

If you have a spare $7,999 (plus a $250 deposit), hate folding laundry, and happen to…

Verify with primary source before acting.

The Verge Tech
News 8h ago

For data scientists, working with high-dimensional data is part of…Choosing Between PCA and t-SNE for Visualization

For data scientists, working with high-dimensional data is part of daily life.

Verify with primary source before acting.

Machine Learning Mastery
Press 8h ago

Anton Cherepanov is always on the lookout for something interestingAI is already making online crimes easier. It could get much worse.

Anton Cherepanov is always on the lookout for something interesting.

Verify with primary source before acting. Also: check API docs for breaking changes;…

MIT Technology Review
Press 8h ago

EVs are getting cheaper and more common all over the worldWhy EVs are gaining ground in Africa

EVs are getting cheaper and more common all over the world.

Verify with primary source before acting. Also: check API docs for breaking changes;…

MIT Technology Review
Press 8h ago

Rather than offering a revolutionary new approach to gig work,…I Tried RentAHuman, Where AI Agents Hired Me to Hype Their AI Startups

Rather than offering a revolutionary new approach to gig work, RentAHuman is filled…

Verify with primary source before acting.

Wired AI
Press 9h ago

MIT Technology Review ’s What’s Next series looks across…What’s next for Chinese open-source AI

MIT Technology Review ’s What’s Next series looks across industries, trends, and…

Verify with primary source before acting. Also: check API docs for breaking changes;…

MIT Technology Review
Research 10h ago

OpenAI is using a "special version" of ChatGPT to hunt down employees…OpenAI reportedly uses a "special version" of ChatGPT to hunt down internal leakers by scanning Slack and email

OpenAI is using a "special version" of ChatGPT to hunt down employees who leak internal…

Find the core claim, method, and released artifacts. Also: check LICENSE and dependencies.

The Decoder
Research 19h ago

Published on February 12, 2026 12:13 AM GMT In this post, I…Research note: A simpler AI timelines model predicts 99% AI R&D automation in ~2032

Published on February 12, 2026 12:13 AM GMT In this post, I describe a simple model for…

Note scope, timelines, compliance requirements. Also: verify benchmark methodology;…

AI Alignment Forum
Business 19h ago

Chinese AI startup Zhupai aka z.ai is back this week with an…z.ai's open source GLM-5 achieves record low hallucination rate and leverages new RL 'slime' technique

Chinese AI startup Zhupai aka z.ai is back this week with an eye-popping new frontier…

Track implications for your stack/vendors. Also: verify benchmark methodology; check…

VentureBeat
Product 19h ago

OpenEnv in Practice: Evaluating Tool-Using Agents in Real-World…OpenEnv in Practice: Evaluating Tool-Using Agents in Real-World Environments

Hugging Face Blog: OpenEnv in Practice: Evaluating Tool-Using Agents in Real-World…

Check docs/changelog for breaking changes. Also: verify benchmark methodology.

Hugging Face Blog
Product 19h ago

Together AI launches Dedicated Container Inference —…Introducing Dedicated Container Inference: Delivering 2.6x faster inference for custom AI models

Together AI launches Dedicated Container Inference — production-grade orchestration for…

Check docs/changelog for breaking changes. Also: note model size and inference…

Together AI Blog
Research yesterday

Interfacial dynamics in two-phase flows govern momentum, heat, and…SurfPhase: 3D Interfacial Dynamics in Two-Phase Flows from Sparse Videos

Interfacial dynamics in two-phase flows govern momentum, heat, and mass transfer, yet…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · Computer Vision
Research yesterday

In this report, we introduce pplx-embed, a family of multilingual…Diffusion-Pretrained Dense and Contextual Embeddings

In this report, we introduce pplx-embed, a family of multilingual embedding models that…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; check…

arXiv · NLP & Language
Research yesterday

Recent advances in robot learning have generated significant…YOR: Your Own Mobile Manipulator for Generalizable Robotics

Recent advances in robot learning have generated significant interest in capable…

Scan abstract → experiments → limitations. Also: calculate cost at your scale.

arXiv · Machine Learning
Research yesterday

Supervised fine-tuning (SFT) on chain-of-thought data is an…Data Repetition Beats Data Scaling in Long-CoT Supervised Fine-Tuning

Supervised fine-tuning (SFT) on chain-of-thought data is an essential post-training…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · NLP & Language
Research yesterday

Preference optimization for diffusion and flow-matching models…Beyond VLM-Based Rewards: Diffusion-Native Latent Reward Modeling

Preference optimization for diffusion and flow-matching models relies on reward…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · Artificial Intelligence
Research yesterday

The Euclidean distance between wavelet scattering transform…SCRAPL: Scattering Transform with Random Paths for Machine Learning

The Euclidean distance between wavelet scattering transform coefficients (known as…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; review…

arXiv · Machine Learning
Research yesterday

Unified Multimodal Models (UMMs) have shown remarkable progress in…GENIUS: Generative Fluid Intelligence Evaluation Suite

Unified Multimodal Models (UMMs) have shown remarkable progress in visual generation.

Scan abstract → experiments → limitations. Also: verify benchmark methodology; check…

arXiv · Artificial Intelligence
Research yesterday

Hierarchical goal-conditioned reinforcement learning (H-GCRL)…Data-Efficient Hierarchical Goal-Conditioned Reinforcement Learning via Normalizing Flows

Hierarchical goal-conditioned reinforcement learning (H-GCRL) provides a powerful…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · Artificial Intelligence
Research yesterday

Projections (or dimensionality reduction) methods $P$ aim to map…LCIP: Loss-Controlled Inverse Projection of High-Dimensional Image Data

Projections (or dimensionality reduction) methods $P$ aim to map high-dimensional data…

Scan abstract → experiments → limitations. Also: review data licensing and biases.

arXiv · Machine Learning
Research yesterday

Tabular foundation models, such as TabPFNv2 and TabICL, have…TabICLv2: A better, faster, scalable, and open tabular foundation model

Tabular foundation models, such as TabPFNv2 and TabICL, have recently dethroned…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; check…

arXiv · Machine Learning
Research yesterday

The prevailing paradigm in large language model (LLM) development…Weight Decay Improves Language Model Plasticity

The prevailing paradigm in large language model (LLM) development is to pretrain a base…

Scan abstract → experiments → limitations. Also: check API docs for breaking changes;…

arXiv · NLP & Language
Research yesterday

As LLM-based agents increasingly operate in high-stakes domains…FormalJudge: A Neuro-Symbolic Paradigm for Agentic Oversight

As LLM-based agents increasingly operate in high-stakes domains with real-world…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · Artificial Intelligence
Research yesterday

Diffusion language models generate text through iterative…Just on Time: Token-Level Early Stopping for Diffusion Language Models

Diffusion language models generate text through iterative refinement, a process that is…

Scan abstract → experiments → limitations. Also: verify benchmark methodology; note…

arXiv · NLP & Language
Research yesterday

Reliable surface completion from sparse point clouds underpins many…From Circuits to Dynamics: Understanding and Stabilizing Failure in 3D Diffusion Transformers

Reliable surface completion from sparse point clouds underpins many applications…

Scan abstract → experiments → limitations. Also: note model size and inference…

arXiv · Computer Vision
Research yesterday

Reinforcement learning with verifiable rewards has driven recent…Asymmetric Prompt Weighting for Reinforcement Learning with Verifiable Rewards

Reinforcement learning with verifiable rewards has driven recent advances in LLM…

Scan abstract → experiments → limitations. Also: note model size and inference…

arXiv · Machine Learning
/ Search M Mode T Theme