Best read it paired with the LeWorldModel paper.

New work from Chinese researchers - Fast-WAM. A recent findings on World Action Models (WAMs): the core advantage of WAMs is not test-time “imagination” of futures, but the training-time supervision from future video prediction.

Researchers propose Fast-WAM, which makes inference simple, fast, and policy-centric.

Google Introduced TurboQuant a new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency.

NB: On-chip SRAM used better, less HBM needed, and hardware-agnostic (works on Nvidia GPUs too, not just Google TPUs).

Meta partnering with Stripe for checkouts:

- Businesses now sell their stuff directly within Facebook (and later Instagram)
- Turn your Facebook ads into one-click checkouts via a toggle in the stripe dashboard
- Built on the Agentic Commerce Protocol.

Meet Cerebra a multidisciplinary AI board for dementia

Cerebra a team of AI agents for clinicians, boosting their dementia risk assessment accuracy by 17.5%.

Code.

Google presented Vibe Coding XR, a new rapid prototyping workflow that empowers Gemini Canvas w/ the XR Blocks framework to turn user prompts into interactive, physics-aware WebXR applications, allowing creators to quickly test intelligent spatial experiences.

Meet ARM-Thinker the first Agentic multimodal Reward Model that autonomously invokes external tools to ground judgments in verifiable evidence.

Accepted to CVPR 2026.

Integrates 3 multimodal tools:

1. Image Crop & Zoom-in for fine-grained visual inspection.

2.Document Retrieval for multi-page evidence gathering.

3. Instruction-Following Validators for constraint verification.

With a Think-Act-Verify loop, ARM-Thinker can call image crop & zoom-in, document retrieval, and instruction-following validators for evidence-based evaluation.

Built on Qwen2.5-VL-7B with SFT + two-stage GRPO, ARM-Thinker improves multimodal reward modeling, tool-use reasoning, and multimodal math/logical reasoning.

Also introduced ARMBench-VL, a multimodal reward benchmark that requires tool use.

Code

Dataset

Evaluation

Meta introduced TRIBE v2 (Trimodal Brain Encoder), a foundation model trained to predict how the human brain responds to almost any sight or sound.

TRIBE v2 draws on 500+ hours of fMRI recordings from 700+ people to create a digital twin of neural activity and enable zero-shot predictions for new subjects, languages, and tasks.

From the paper:

'The present results strengthen the possibility of a paradigm shift in neuroscience ... moving from the fragmented mapping of isolated cognitive tasks toward the use of unified, predictive foundation models of brain and cognitive functions by aligning the representations of Al systems to those of the human brain, we demonstrate that a single architecture can integrate a vast range of fMRI responses across hundreds of individuals, extending the framework that led the 2025 Algonauts competition.

The observed log-linear scaling of encoding accuracy mirroring power laws in both artificial intelligence and neuroscience suggests that the ceiling for predicting human brain activity is yet to be reached.'

Google shipped Gemini 3.1 Flash Live

It’s built for production-ready reliability.

If you're building agents that need to execute complex tasks at scale, give it a spin.

Stripe launched Projects to help agents instantly provision services from the CLI.

For example, simply run:

$ stripe projects add posthog/analytics

And it'll create a PostHog account, get an API key, and (as needed) set up billing.

Projects is launching today as a developer preview. You can register for access (it available to everyone soon) at projects.dev.

Also rolling out support for many new providers over the coming weeks.

Claude Code can now auto-fix your PR in the background.

All you have to do is turn on the Auto Fix setting and go touch grass.

Cool work by Chroma training a search agent with SoTA efficiency.

Chroma Context-1, a 20B parameter search agent:

•pushes the pareto frontier of agentic search

•order of magnitude faster

•order of magnitude cheaper

•Apache 2.0, open-source

Lots of cool details: a prune tool for editing context mid-search, a synthetic data pipeline with verification steps, and a curriculum that shifts from recall to precision.

Trained with Tinker.

Stanford proposed a bold new vision for biology: Virtual Embryos

By integrating single-cell and spatial genomics with AI, researchers can build digital twins of embryogenesis—moving beyond virtual cells to predict cell growth, division, migration, state transitions, and morphogenesis, from genes → cells → organs -> organ systems and whole embryo, in a fully 4D spatiotemporal framework.

This approach could enable truly predictive biology and in silico experimentation to diagnose, prevent, and treat developmental diseases transforming medicine and improving outcomes for future generations.

Unitree open-sourced UnifoLM-WBT-Dataset a high-quality real-world humanoid robot whole-body teleoperation (WBT) dataset for open environments.

Publicly available since March 5, 2026, the dataset will continue to receive high-frequency rolling updates. It aims to establish the most comprehensive real-world humanoid robot dataset in terms of scenario coverage, task complexity, and manipulation diversity.

Nicolas Carlini at Anthopic just showed Claude finding zero day vulnerabilities in a live conference demo

Claude has found zero day in Ghost, 50,000 stars on github, never had a critical security vulnerability in its entire, history...

it found the blind SQL injection in 90 minutes, stole the admin api key, then did the exact, same thing to the linux kernel.

HuggingFace built infra where an agent can launch GPU experiments, access all of hf datasets as local filesystem, and semantically search arxiv for ideas.

Mathematical methods and human thought in the age of AI by Tanya Klowden, Terence Tao

In this paper, researchers consider the rapidly evolving impact of AI to the traditional questions of philosophy with an emphasis on its application in mathematics and on the broader real-world outcomes of its more general use.

They are assert that AI is a natural evolution of human tools developed throughout history to facilitate the creation, organization, and dissemination of ideas, and argue that it is paramount that the development and application of AI remain fundamentally human-centered.

They are propose a pathway to integrating AI into our most challenging and intellectually rigorous fields to the benefit of all humankind.

Microsoft Introduced Critique, a new multi-model deep research system in M365 Copilot.

You can use multiple models together to generate optimal responses and reports.

New supply chain attack this time for npm axios, the most popular HTTP client library with 300M weekly downloads.

It's possible to personally defend against these to some extent with local settings e.g. release-age constraints, or containers or etc, but ultimately the defaults of package management projects (pip, npm etc) have to change so that a single infection (usually luckily fairly temporary in nature due to security scanning) does not spread through users at random and at scale via unpinned dependencies.

Supply chain attacks like the currently breaking axios, litellm and xz are only going to be more commonplace in the vibecoding world.

The entire premise of vibecoding is “I don’t need to understand the code” happens to also be the entire premise of a supply chain attack.

Claude code source code has been leaked via a map file in their npm registry. You can find it here.

OMG! Whoop raised $575M at a $10.1B valuation

Whoop is building the world’s leading personal, preventive health platform powered by continuous biometric data, advanced analytics, and AI to help people understand their bodies and improve their health in real time.

In the past 12 months, Whoop has received medical clearances, launched blood testing, and created a platform that has saved lives.

Abbott and Mayo Clinic two of the most respected and influential institutions in global healthcare are now investors in Whoop.

Google presented new paper on AI Agent Traps

An increasing volume of web content is being created by, and consumed by, advanced AI agents.

This puts environmental AI safety in focus, as it exposes a vast attack surface via the content that AI agents interact with.

This paper explores the landscape of environmental attacks and defenses, aiming to inform mitigations that are needed for ensuring safety of the agentic web.