Eric Yates

@ericyates.bsky.social

134 Followers 137 Following

On a mission to reveal the causal mechanisms of unexplained neurological conditions, starting with my own I’ve built robust software for real-time biosignals over 5 years of SBIR/STTR grants Functional neuroanatomy + insula 🧠🤖💜👋🏻 github.com/neurodevs

Screenshot of a Notion database table with Type, Provider, and Name columns, listing: Courses: Medea & Other Plays (Euripides) (Memoria Press), Ancient Greek Civilization (The Great Courses), Greek Tragedy (The Great Courses), Introduction to Greek Philosophy (The Great Courses), Masters of Greek Thought: Plato, Socrates, and Aristotle (The Great Courses), The Long Shadow of the Ancient Greek World (The Great Courses) Documentary: Greece: Secrets of the Past (IMAX), Alexander's Lost World (Athena), 5000 Years of Magnificent Wonders: Ancient Greece (Questar), Ancient Civilizations: Athens and Ancient Greece (Questar), Ancient Greece: Gods and Battles (The History Channel), In Search of History: The Greek Gods (The History Channel), The True Story of Alexander the Great (The History Channel), In Search of the Trojan War (BBC), Helen of Troy (PBS), Odysseus Returns (PBS), The Greeks: Crucible of Civilization (PBS), The Real Olympics (PBS)
There are various biosensors shown, including all 6 current models of the Muse headset: Muse 1 Gen 1, Muse 1 Gen 2, Muse 2, Muse S Gen 1, Muse S Gen 2, Muse S Athena. There are a few other biosensors shown that are not directly relevant to this post.
Screen recording of Visual Studio Code showing the libndx project. The file explorer on the left lists the project's directory structure (include/ndx, src, tests, etc.), and the editor shows usb_provider.cpp with a PosixUsbProvider class. A terminal panel on the right streams real-time data, with each line printing a sample index and readings for 8 channels (ch1–ch8).
The OpenBCI 8-channel Cyton Biosensing Board is shown, with its variety of parts laid out in an orderly fashion. At the center of the image, there is a box that has its electrodes and cables shown, along with an OpenBCI sticker on the front left of the box. In front of the box, there is a paper card that shows the contents of the box, with both a name and a visual representation for each part. The parts themselves are laid around the card.
A “biosensor tower” that is a rounded piece of cardboard with 4 biosensors running from top to bottom: Muse 2, Muse S Athena, Muse S Gen 2, and Zephyr BioHarness 3.0.
Screenshot of a webpage footer displaying a GitHub-style badge labeled "Neurodevs" with "22 repos, 540 tests" in green, positioned below a symmetrical diagram of brain regions labeled left (L) and right (R).
Terminal screenshot showing a script output for test counts across @neurodevs repos. The script, executed with node build/scripts/demo.js, displays a JSON-like object: total: 536, with a perRepo breakdown. The repo with the highest count is 'react-connectivity-graphs': 78, followed by 'node-ble': 71 and 'node-lsl': 70. Repos with the fewest tests include and 'node-test-counter': 3, 'node-html-loader': 2, and 'node-eeg': 2.
Robotic arm begins in a vertical orientation, then performs a two-stage motion: first sweeping downward and outward in a wide arc, followed by a brief pause, then returning to vertical along a shorter, more direct path.
A terminal window displays output from running a Node.js script (node build/scripts/demo.js) for a causal model used in the I-INSULA protocol. The causal model M is shown as a triple ⟨U, V, F⟩, where U is a set of external variables (like C-tactile, vestibular, thermal, and visual stimuli), V is a set of internal biosignal variables (like source-localized EEG and EMG), and F is a set of causal mechanisms represented as functions. A submodel Mx is also shown, where one function in F has been modified to return a constant value (e.g., 1) to simulate activity of an internal variable (in this example, posterior insular cortex activity) to test whether it is sufficient to causally explain downstream activity in the cingulate cortex or EMG muscle activity.
Screenshot of a TypeScript class named StructuralCausalModel implementing Judea Pearl’s formalism of a structural causal model. The class defines private readonly fields U, V, and F for external (exogenous) variables, internal (endogenous) variables, and causal mechanisms, respectively. It includes a static Create method, a toTriple method, and supporting interfaces for typing the causal model components. The code structure reflects a modular, type-safe design for executable causal modeling.
A screen recording shows a 3-column grid of 9 side-by-side circular flow graphs, each containing a different number of nodes from 1 to 9. The graphs are rendered in real-time using a consistent visual style. The layout allows for rapid visual comparison of spacing, label wrapping, and node positioning as the number of nodes increases. The demonstration highlights efficient testing of visual consistency across variations.
A video demonstrating a flow graph with distinct left and right sides. As the cursor hovers over each node on the left, it reveals the full label and all connected nodes/edges. Each node has a fixed width, so longer labels automatically wrap onto multiple lines for visual compactness.
An image showing git changes in TypeScript code, with 2 changes made. A minus sign was changed to a plus sign, which fixed a known issue. Text overlay on the image also describes the thought process.
A video showing a personal website with attention drawn to the new footer, displaying: © 2025 by @ericthecurious. Licensed under CC BY 4.0.
A video demo of a minimal personal website featuring a single post titled Hello world! The post displays the text: “This is an MDX file. It can contain both Markdown and JSX components, like this.” The cursor then moves across various regions of an interactive flow chart embedded in the post, highlighting the responsiveness of the visualization to user interaction.
Short, looping video showing a menu being open and closed. The menu fades and slides in while opening, and it fades and slides out while closing.
A video showing a user click on a menu button which turns blue and opens a menu below with the menu items Posts, Projects, and About. When the video repeats itself from looping, it then appears like the user clicked the menu button again to turn it dark grey, and then the menu closed and showed the original content of Hello world!
A minimalistic webpage with a header and “Hello world!” below it. The header includes a logo with a blue square and a blue question mark within the square, @ericthecurious, Posts, Projects, and About. The items are visually styled to include ample space around them.
A very minimal webpage for a personal website. A header shows @ericthecurious, Posts, Projects, and About. The main body shows “Hello world!” The rest is blank space.
A flowchart illustrates the neural pathway beginning with a cursor on a left-sided C-tactile receptor in the forearm. The pathway travels to the cervical spinal cord lamina I / II. From there, the cursor crosses over to the right-side ventromedial posterior thalamic nucleus. The pathway then projects to the posterior insular cortex, which relays it forward to the ventral anterior insular cortex. From there, the cursor continues to the posterior midcingulate cortex, which in turn connects to the primary motor cortex. The final output splits into three branches, each projecting to a different lower motor region.
Animated lateral flow graph showing a neural pathway that starts on the left side of the body and continues on the right. It begins with left C-tactile afferents, then passes through left dorsal root ganglion, left spinal lamina I/II, right ventral medial posterior nucleus, right dorsal posterior insula, right middle insula, and ends in right anterior insula. Hovered nodes and their direct connections appear in color with full names, while unrelated nodes stay dimmed and show only abbreviations.
Flow chart showing a lateralized structure divided by a vertical green line. The left side contains nodes representing the left side of nervous system structures, while the right side mirrors complementary right-sided structures. The green line visually emphasizes the division between the sides in the system. Nodes are connected by directional edges indicating flow, forming a symmetric yet distinctly bifurcated pattern.