The papers
Each published discovery is a real paper: one dataset, one falsifiable question, one measured answer. Open it in the browser or download the arXiv-grade PDF, typeset from the same record. Negative results are published here too.
2 papers published
A predictor that sees zero pixels and knows only where a frame sits in time reaches 70.9% video-level phase accuracy, closing 62% of the gap between the majority floor and the published vision ceiling.
A small MLP on fused RDKit descriptors and count fingerprints matches Chemprop's D-MPNN on both scaffold and random ESOL splits, with overlapping 95% confidence intervals, while using 3.6 to 4.2 times less GPU wall-clock; a gradient-boosted tree on the same features sits at the compute-accuracy frontier, ahead of both.