You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Dandelion Engineering research on whether distributed structural sensing helps simulated robots detect and adapt to body and sensor changes. Ended 2026-08-23 with no research question answered — because of the framework it ran on, not the question. Kept public as a record; successor: Collaboration Station Research 2.0.
A worked demonstration of the Collaboration Station™ framework running an AccuSleePy sleep-staging pipeline end to end, with a reproducible pipeline and a written report. A demonstration, not a Dandelion research project.
A worked demonstration of the Collaboration Station™ framework running a 4,000-file multi-condition vibration batch analysis, with a reproducible pipeline and a technical report. A demonstration, not a Dandelion research project.
The general Collaboration Station™ framework for long-running human–AI collaboration: persistent agent memory, asynchronous multi-agent work, and a file-based audit trail of every decision. Dandelion's exploratory research projects run on the research-specific Collaboration-Station-Research.
A filesystem-native template for long-running exploratory research by human researchers and AI agents, rebuilt for current self-verifying models. Early testing.
A filesystem-native framework for humans and AI agents to run long-lived computer-based projects with durable context and lightweight coordination. Early testing.
Open Dandelion Engineering research asking whether scalp EEG carries a subject-transferable signature of memory load in the medial temporal lobe. A pre-registered negative result, shipped with the packet to check it.
Dandelion Engineering research on whether the realism of hybrid ground truth changes measured spike-sorting accuracy, and the ranking between sorters. Ended 2026-08-23 having never taken a measurement — because of the framework it ran on, not the question. Kept public as a record; successor: Collaboration Station Research 2.0.