Supervised Feed

Artificial Intelligence, Internet of Things, Cloud and more. #MachineLearning #AR #VR #Automation

[Supervised Feed @FeedSupervised]

Digital twins are virtual models of physical entities used to optimize business operations. Success lies in understanding best practices, defining business goals, testing via prototyping, assessing organizational readiness, and adopting an agile approach.

RT @antgrasso

[PyTorch @PyTorch]

This is your final opportunity to submit a session for #PyTorchDayChina taking place in Beijing on December 12. Bring your ideas and experiences to a community of PyTorch developers, machine learning engineers, AI researchers, and open source AI leaders from across China and beyond.

⏰ SUBMIT YOUR PROPOSAL by Tuesday, November 10, at 23:59 China Standard Time >> https://t.co/szlAXuZiFV

[Django @djangoproject]

Django is grateful to be a recipient of part of Sentry's generous Open Source pledge. Sentry is giving a total of one MILLION to open source maintainers.

[Kirk Borne @KirkDBorne]

Probability and Statistics Essentials for Data Science and Machine Learning, with 200+ examples and illustrations: https://t.co/Q4OXxax0yI

[AI at Meta @AIatMeta]

During Connect, we previewed Muse Realtime Avatar, our real-time embodiment technology that turns Muse Realtime Voice into expressive, interactive avatars.

Today, we’re rolling it out in the @Muse app in the US, giving people a more natural way to engage with their personal agents.

Learn more about Muse Realtime Avatar on our research blog: https://t.co/V2cSOHunRw

[CEO Tech Awareness @CeoAwareness]

Leadership loses depth when speed becomes the only metric. Sound decisions require space for reflection, attention to others, and the patience to let ideas mature. When we slow our pace, we strengthen trust and create conditions where teams can progress with confidence.

[Antonio Grasso @antgrasso]

Ripple builds payment infrastructure for cross-border transfers.

XRP can serve as a bridge asset on the XRP Ledger, which settles in seconds with low network fees.

The point is simple: move value across systems without relying only on pre-funded accounts.

Microblog @antgrasso

[DataCamp @DataCamp]

Last hours of Free Access Week. Finish a course today and leave with a working skill.

750+ courses, tracks, and certifications, free to access until 11:59 PM ET tonight.

[Kirk Borne @KirkDBorne]

Understanding Deep Learning: Building Machine Learning Systems with PyTorch and TensorFlow: From Neural Networks (CNN, DNN, GNN, RNN, ANN, LSTM, GAN) to Natural Language Processing (NLP)

Get it here: https://t.co/VfF1wF4vOj

[Supervised Feed @FeedSupervised]

We often treat automation as a way to lighten our load, but real innovation is never about washing our hands of responsibility.

When data and models guide decisions, people still need to know who approves them and who answers for the outcome.

By @antgrasso #HumanAI.

[Kirk Borne @KirkDBorne]

Reinforcement Learning — Industrial Applications of Intelligent Agents: https://t.co/1eoKUMwp5p

Practical book with these benefits:
🔴 Learn what RL is and how the algorithms help solve problems
🔴 Become grounded in RL fundamentals including Markov decision processes, dynamic programming, and temporal difference learning
🔴 Dive deep into a range of value and policy gradient methods
🔴 Apply advanced RL solutions such as meta learning, hierarchical learning, multi-agent, and imitation learning
🔴 Understand cutting-edge deep RL algorithms including Rainbow, PPO, TD3, SAC, and more
🔴 Get practical examples through the accompanying website

[IBM Research @IBMResearch]

The @QuantumSciCtr at @ORNL, @IBM, and @AMD deploy a new co-design platform for quantum error correction at IBM’s site in Poughkeepsie, New York.

The platform enables low-latency quantum-classical workflows for advancing quantum computing research: https://t.co/tCI0WKswba

[DataCamp @DataCamp]

Start building:
https://t.co/CIv5WguxHf

[Anaconda @anacondainc]

Every AI agent introduces decisions that teams may need to explain and audit.

This workshop explores risk classification, automated guardrails, and audit trails, with a live Enkrypt demo showing policy enforcement in action.

Watch the free workshop: https://t.co/xx1PSfRgco

[Anaconda @anacondainc]

See Anaconda CLI + MCP in action with Kilo Code.

We walk through setup, then use natural language to list conda environments, create one with Python 3.12, install packages from a secure channel, and run K-means clustering.

Watch the demo: https://t.co/HaL4WtfeU4

[DataCamp @DataCamp]

Paying a reasoning model to label a support ticket is paying for an essay when you wanted one word.

Decision models return one answer with a confidence score and hand the unsure cases to a reasoning model. Our top 10 for 2026 is in the reply below.

[Supervised Feed @FeedSupervised]

Wind farms, consisting of multiple turbines, efficiently convert kinetic energy into clean, renewable power, helping to reduce greenhouse gas emissions and optimize energy production.

#RenewableEnergy #Sustainability #GreenTech.

[Antonio Grasso @antgrasso]

An "Access denied" message does not always mean something went wrong. Sometimes the system knows exactly who you are, and still says no.

Why should a successful login open every application to every employee?

Access permissions exist for a reason.

Microblog by @antgrasso

Authentication verifies your identity. Authorization determines which applications and information you are allowed to access.

Consider an employee who moves to another department. The company account remains valid, but some permissions associated with the previous role may no longer be appropriate.

Those permissions should change with the work.

And identity is not always the only condition. For sensitive applications, security policies may also check whether the device meets company requirements before allowing access.

So an employee can be correctly authenticated and still be denied access to a particular resource.

There is nothing contradictory about that.

People can share the same corporate network without having the same permissions. Access should reflect what each person is authorized to do, even as responsibilities change.

#IdentitySecurity #AccessControl

[AI at Meta @AIatMeta]

Today we're bringing specialized models trained using Muse Wonder to @Muse and Meta AI, giving users access to some of Muse Wonder's frontier capabilities for complex tasks at work and in everyday life.

These specialized models include Muse Wonder's Active Memory and Active Focus, helping them stay aligned with your goals as tasks unfold:

→ Active Memory helps recognize when earlier information matters and distinguish what's still relevant from what's become stale

→ Active Focus helps recover from setbacks or find alternatives when constraints change

In Muse, the improved models show a 2.4x reduction in forgetting information critical to your goals and a 67% improvement in getting unstuck as constraints evolve. That means less repeating yourself and fewer dead ends when plans change.

In Meta AI, these improvements help it break complex problems into steps, ask sharper follow-up questions and adapt when you change direction.

Read more on our research blog: https://t.co/ACWFr9xPtS

[DataCamp @DataCamp]

Spend your weekend on a skill that pays off on Monday.

Last days for Free Access. 750+ courses and tracks on today's top technologies: Claude, SQL, Power BI, Python, Microsoft Copilot, and more. All free through Sunday.

[Thomas Wolf @Thom_Wolf]

18 of the 20 best open models today are Mixture-of-Experts. Dense is basically over.

And yet the know-how to train a good MoE still lives mostly in the heads of a handful of teams.

So we spent 800k GPU hours running the experiments ourselves and wrote everything down. Super happy to finally release "The Ultimate MoE Training Handbook", free and open, organized around the questions training teams hit:
- how many experts, how big, shared or not?
- why is half my step spent in all-to-all, and how do I hide it?
- why is my run 25x slower with no error in the logs?
- softmax, sigmoid or sqrt(softplus) router, and why keep it in fp32?
- aux loss or DeepSeek's loss-free bias?
- upcycle a dense checkpoint or train from scratch?
- can I serve with fewer experts per token than I trained with?

⚠️ Warning: it's long (~70k words, budget a few days, sorry) but with 55 interactive figures, code you can run and numbers from our own runs.

Read it here: https://t.co/6XihlUrj92