RED-2400: A Public Benchmark of Algorithmically-Rejected Trading Events with Outcome Labels

RED-2400 is a public benchmark of 6,660 algorithmically-rejected trading events drawn from a live Solana decentralised-exchange filter stack, observed continuously between 2026-04-10T21:10Z and 2026-05-02T21:48Z (UTC), 22 calendar days. Each rejection event is linked to its post-rejection price-and-liquidity trajectory: 169,123 forward-outcome observations across 1,837 graveyard-tracker lifecycle snapshots. Outcome labels follow the five-tier classification rule of the related methodology paper (Kamat 2026c, SSRN abstract_id 6638259, separate Zenodo deposit at concept DOI 10.5281/zenodo.19987695): saved (windowed), saved (early-death), missed, flat, unclassifiable. The deposit's lifecycle-tracker file (graveyard_lifecycle.csv) permits external validation of the rule against observed token-lifecycle ground truth; an external validation is reported in §V of the companion paper. Filter labels are anonymised to filter_1 through filter_8 in descending order of rejection volume. Source-collector identifiers are anonymised to source_a and source_b. Liquidity and 24-hour volume are log2-binned to preserve heavy-tailed distributional shape while blocking threshold inference. Filter-internal model outputs (signal scores, weights, calibration thresholds) and the priceChange* derivatives are removed. The graveyard to-state vocabulary in the file is exactly: alive_active, alive_dormant, gone. SHA-256 checksums for the three CSV files are provided in checksums.txt and are byte-identical between all versions of this deposit. "RED-2400" is a brand name; the number does not equal any cohort size in the current deposit. This is the first window of a planned dataset series; subsequent windows will extend the time horizon and enable regime-stratified analysis.