🛸Mayank Sharma🏴‍☠️

TISO | AI Platform Security Engineer/Architect @DeutscheBank 💶 | Crew @Cloudvillage_dc 🌩️| Hack Ship Hack☁️ 🐳📦 Cloud Cotainers

[JFrog @jfrog]

Thank you, London! What a week at @Infosecurity Europe 🐸

Hundreds of conversations, a packed Case Studies Theatre session with Ofri Ouzan talking about securing AI agents, and some brilliant questions from security leaders who are taking the software supply chain seriously.

The AI-era security conversation is just getting started.

Missed us at the event? The team is always available. Visit the link below to continue the conversation 👇
https://t.co/u234fPWRqQ

#DevSecOps #SoftwareSecurity #InfoSecEurope

[0xSero @0xSero]

This guy is worth following he’s fuel for motivation.

[Atai Barkai @ataiiam]

All UI will become AI

This is THE article you'll need give your agents a frontend.

Agent-user collaboration is the future, see below 👇

[Vesa Juvonen @vesajuvonen]

✨ Introducing Microsoft Scout: Your always-on personal agent

• A new category of agents called Autopilots
• Introducing Microsoft Scout
• Integrated across the Microsoft 365 apps you use every day
• Enterprise compliancy and control

See more → https://t.co/VnNwnx1NDX

#Copilot #Microsoft365

[Omar Shahine @OmarShahine]

Introducing Microsoft Scout, the first autopilot agent from Microsoft - 57 days after starting my new job, we are launching Microsoft Scout to our Frontier customers. Big day for the team and for @openclaw #MicrosoftBuild https://t.co/ZGl47ADnCg

[Google Open Source @GoogleOSS]

Building the trust layer for AI is a team effort. 🌉 Join Google at @linuxfoundation's Confidential Computing Summit in SF, June 23-24! Don't miss Nelly Porter's opening Keynote on 6/23 at 9 AM. How are you scaling AI workloads securely? 🚀 #CCSummit https://t.co/bNFn07odaJ

[fwd:cloudsec @fwdcloudsec]

The conference has started! 🎉

[fwd:cloudsec @fwdcloudsec]

It's conference day! Check out the schedule and livestream links below. https://t.co/lUTl8rE6Fc

[Perplexity @perplexity_ai]

More info about Search as Code in the Perplexity Agent API docs:

https://t.co/iaXcFSmiWh

[Perplexity @perplexity_ai]

Personal Computer is now available to all users in a new Perplexity Mac app.

Personal Computer is an advanced version of Perplexity Computer.

It operates on any Mac, running tasks across your local files, native Mac apps, the web, and Perplexity’s secure servers.

[Perplexity @perplexity_ai]

Introducing Search as Code, our new search architecture for AI agents.

It writes Python that calls our search stack directly, instead of looping through function calls one at a time.

Available in the Perplexity Agent API, and now default in Computer.

https://t.co/ut6GGWQTVO

[Aravind Srinivas @AravSrinivas]

We’re moving away from search as a web fetch tool call to search as codegen to be future proof in a world where code execution inside agent harnesses is the way to do almost all of our knowledge work.

Doing this lets you compose multi-step primitives far more naturally and be much more adaptable to changes made to the agent harness, as well as benefit from improvements in coding capabilities that are guaranteed to come from the next generation of frontier models.

[Adobe @Adobe]

Adobe is partnering with @nvidia to optimize creative workflows across apps like Photoshop, Premiere, and Substance 3D with @NVIDIARTXSpark, helping creators spend less time waiting and more time bringing their ideas to life. From faster editing and effects to more responsive 3D workflows, this is about giving creatives more power, precision, and flow across the tools they use every day.

Learn more: https://t.co/3BoZDc0fvr

[NVIDIA RTX Spark @NVIDIARTXSpark]

NVIDIA RTX Spark reinvents @Windows PCs for the era of personal AI agents, offering a new class of computer that moves from tool to teammate.

[NVIDIA RTX Spark @NVIDIARTXSpark]

Full blog: https://t.co/8tDsu8YWzd

[Theo - t3.gg @theo]

swe-bench is kind of a shitshow, and it makes evaluating LLMs hard. DeepSWE is the first agentic code bench that makes sense.

[Dark Web Informer @DarkWebInformer]

🚨 Security researchers are now handing over vulnerabilities to Nightmare Eclipse after he was banned on both GitLab and GitHub.

It should be a fun month, because man has it been boring the last couple of weeks.

[🛸Mayank Sharma🏴‍☠️ @ping_mayank]

Another prompt injection lead to an RCE in antigravity https://t.co/epRKRdzCAa

[Roy van Rijn @royvanrijn]

For curious developers 🧠

I built "The Anatomy of an LLM", an interactive explainer showing how text becomes tokens, vectors, attention, transformer blocks, and finally generated text.

https://t.co/fgCeZuQwJf

[Pavlo Molchanov @PavloMolchanov]

What if you could take three completely different model families… and distill them into one tiny model? 🤯

📜 Paper: https://t.co/K2iKD4xFvp

MOPD (Multi-Teacher On-Policy Distillation) has become a standard procedure in post-training. We already distill multiple specialized variants of the same model into a single set of weights.
But what if we could go further - and distill models from entirely different families? Turns out, it is possible.

Today we’re releasing a paper on cross-tokenizer distillation - our first steps in this exciting direction. 📄

We distilled Qwen3-4B, Phi-4-Mini, and Llama-3B into Llama-3.2-1B.
MMLU jumped from 32.05 → 46.32 when using multiple teachers. 📈

The team is now working on Nemo-RL integration so the community can try this method in their own settings. Plus, we are scaling experiments up. 🚀