Skip to content
No results
  • Frontier Radar
  • Courses
  • Open Source Projects
  • Engineering
  • English
    • 中文
    • 日本語
NoteFlowAI brand logo
Physical AI Lab
  • Frontier Radar
  • Courses
  • Open Source Projects
  • Engineering
  • English
    • 中文
    • 日本語
Sign in
NoteFlowAI brand logo
Physical AI Lab
Sign in
Replanning cycles, early exit, fewer denoising steps and quantization appear as parallel topics without a causal or execution order. A central clock names latency, and the quantization note states that saving memory does not imply speed, with no new measured values.
  • Frontier Radar

Frontier Radar deep dive | Edge and real-time: where the latency budget of VLAs and world action models goes

9 sources: Rolling-WAM cuts replanning from 978 to 215 ms, early exits plus fewer denoising steps cut ~80% of latency; quantization saves memory, not time.

  • September 28, 2026
Four learning formats—video, interactive pages, notebook and quiz—shown as alternatives without a required order.
  • Courses

Physical AI Lab: a physical AI course where every number can be checked

Updated 2026-10-13: 15 lessons available in Chinese, English and Japanese, with videos and sources. In-depth reading and exercises require sign-in.

  • September 28, 2026
Course cover · 0.2
  • Courses

Physical AI Lab 0.2: Why Now? The Three Curves of Data, Models, and Compute

Why Now? The Three Curves of Data, Models, and Compute. English video, reading, self-checks and sources in one lesson, updated 2026-10-04.

  • September 28, 2026
Course cover · 0.1
  • Courses

Physical AI Lab 0.1: What Is Physical AI? From Chatting to Doing

What Is Physical AI? From Chatting to Doing. English video, reading, self-checks and sources in one lesson, updated 2026-10-03.

  • September 28, 2026
Original concept diagram: simulation records branch into browser replay, raw data labelled for SHA-256 verification, and editable Blender/OpenUSD scenes, without third-party simulation-asset screenshots.
  • Open Source Projects

Robot Reel: turning robot policy simulations into frame-by-frame verifiable replays and 3D scenes

Turns sim runs into browser replays, hash-verified raw data and editable Blender/OpenUSD scenes; core: a paired stress test of 30 closed-loop SmolVLA runs.

  • September 28, 2026
Radar arcs symbolize discovery; O, R and M label official first-hand sources, research preprints and secondary media, with an up-to-8 daily limit and eight lanes rather than a daily item count.
  • Open Source Projects

Physical AI Radar: an embodied AI morning brief with an evidence level on every item

Up to 8 embodied AI items a day from arXiv, vendor blogs and media, in eight lanes, tagged O/R/M, numbers quoted as published; open rules, reproducible offline.

  • September 28, 2026
Two run documents labelled A and B connect to a central difference-inspection lens, with a separate empty-box acceptance-rule checklist below; no passing verdict is implied.
  • Open Source Projects

EvalArc: the score went up, so where did the agent regress?

Compares two agent eval runs check by check under explicit acceptance rules to catch regressions hidden by a higher average, and audits graders on known bugs.

  • September 28, 2026
Multiple SKILL.md documents from tools, plugins and Git converge on one catalog; SHA-256 labels file verification before on-demand loading through dsh or MCP.
  • Open Source Projects

Skills Anywhere: one catalog for the SKILL.md files spread across AI tools, loaded on demand through dsh or MCP

Merges Agent Skills from Claude Code, Codex, Cursor, plugin marketplaces and Git into one catalog, served via DeepSeek Harness or MCP, with file/hash checks.

  • September 28, 2026
A central WordPress chat interface connects to Amazon Bedrock; the bidirectional MCP connection represents the plugin’s server and client roles.
  • Open Source Projects

AI Agents & Chat for Amazon Bedrock: the WordPress chat and MCP plugin this site runs on

Open-source WordPress plugin for Amazon Bedrock on your own AWS account: streaming chat, a controlled agent, MCP server and client, rule-based answer checks.

  • September 28, 2026
Version consistency, release inspection and SVN appear on the left; a separate right-hand note links same-version re-upload to stale CloudFront CSS, not to SVN.
  • Engineering

Lessons from releasing a WordPress plugin: version numbers, the release gate, SVN and CloudFront caching

Releasing AI Chat for Amazon Bedrock: 3-way version check, a gate that reports skip, SVN, and CloudFront's stale CSS by ?ver= after a same-version re-upload.

  • September 28, 2026
Three layers show CloudFront/ALB routing, the WordPress application, and persistent EFS/Aurora state.
  • Engineering

Hosting WordPress on AWS with CDK: trade-offs of a shared ALB, CloudFront, EFS and Aurora

This site's WordPress on AWS: Fargate, the site on EFS, Aurora Serverless v2 pausing to 0 ACU, a shared ALB gated by a CloudFront secret header. Choices, costs.

  • September 28, 2026
A flat viewer–CloudFront–ALB–Apache–PHP chain contrasts viewer-influenced X-Forwarded-For with the viewer-address header generated by CloudFront.
  • Engineering

Getting the visitor IP behind CloudFront + ALB: why you should not parse X-Forwarded-For in PHP

The WordPress image's mod_remoteip already rewrites X-Forwarded-For, so parsing it in PHP yields forged IPs. Use CloudFront-Viewer-Address, given 2 conditions.

  • September 28, 2026
Prev
1 2 3 4 5 6
Next
© 2026 Physical AI Lab
  • Privacy Policy

Physical AI Lab · AI Assistant

AI
Hi! I'm the Physical AI Lab's AI assistant. Ask me anything about robotics, embodied AI or the course.

Sign in to chat with the assistant.

Sign in
Powered by Amazon Bedrock