Zyphra just dropped ZUNA — the first foundation model built specifically for raw EEG data. Trained on ~2 million channel-hours of scalp EEG, it can denoise noisy signals, reconstruct missing channels, and even predict what electrodes you don't have would be reading using 4D positional encoding of electrode coordinates. It works across different headset configurations out of the box, which means it could slot right into an 8-channel Cerelog setup.

The request: We'd love to see someone try piping live Cerelog ESP-EEG data through ZUNA and pushing the decoded output toward thought-to-speech. The model is already positioning itself as a step toward thought-to-text — pairing it with our board could be one of the first real-world attempts at noninvasive thought-to-speech with affordable, open hardware. If you're into BCI, ML, and want to attempt something genuinely novel — this is it.

A browser-based playground for building and testing BCI decoders. A decoder is a small program — usually a machine-learning model — that looks at those signals and guesses what the person was thinking or trying to do. Streams live data from your Cerelog ESP-EEG, runs a decoder on it, and plots predictions against ground truth side by side. Bring your own decoder in JavaScript or TensorFlow.js, or generate one from a text prompt.

A brain-computer interface game that lets players control Pong paddles using eye movements. Uses electrooculography (EOG) to detect electrical signals from eyeball movement, processes them through a multi-stage signal processing pipeline (filtering, detrending, spike detection), and translates them into real-time game controls. Built with the Cerelog 8-channel BCI board and a Python-based processing pipeline — successfully tested in a competitive tournament with multiple participants.

A wearable EEG headband that holds dry electrodes against the scalp at standard 10–20 positions. Unlike rigid 3D-printed EEG caps, the electrode array is built from stapled nylon webbing and elastic band segments — so it flexes to fit the head, packs flat, and is cheap to build and repair. Electrodes mount as a small "sandwich" (dry comb electrode, 3D-printed washer, machine screw, and jumper-wire lead) locked with hot glue. Signals are digitized by a Cerelog ESP-based board housed in a 3D-printed case that velcros to the back of the band.

Includes full assembly guide, STL files for the electrode washer, Cerelog case, and LiPo battery holder, plus a complete bill of materials. Work in progress — designs and instructions may still change.

Custom firmware that turns the Cerelog into a confound and timing recorder instead of an EEG board. Four differential channels record jaw muscle activity at gain 24, and a fifth carries an ALS-PT19 light sensor through a simple voltage divider at gain 1 — so muscle artifact and on-screen stimulus timing land in the same data frame at 1,000 SPS. Built for PRAYCG, an open neuroscience workbench. Package includes full source, binaries, a Python reader, and bench checks.

What's PRAYCG? An open, local-first workbench and protocol for studying how the body and brain respond to meaningful stories. It asks a narrow question: when someone watches a coherent narrative, does their physiological state follow a different path than when they watch the same footage scrambled so the meaning is destroyed, or watch it intact but under an analytic task?