Showing posts tagged #jev. Clear
Part 7 of 7 · Jev in trading, tested
With QuantDinger's defaults, the Jev gate decided 0 of 1,324 entries and let every trade through. With the thresholds removed, it did no better than a coin flip.
Part 2 of 2 · Jev on Polymarket
I assumed a model trained on the internet would know how 2025's biggest markets resolved, so a backtest on old Polymarket data would be leaky. Jev scored the same on 17,000 old markets as on unseen ones.
Part 1 of 2 · Jev on Polymarket
I asked Jev "will this resolve Yes?" for 26,174 resolved Polymarket markets. On the 9,138 resolved after its release it scored AUC 0.615. The market price a day earlier scored 0.844. Code included.
Part 6 of 7 · Jev in trading, tested
Jev misread Bitcoin's indicators in my test (AUC 0.486 vs 0.522 for a plain model). Where Jev does help traders, and a 5-question checklist for AI claims.
Part 5 of 7 · Jev in trading, tested
I gave Jev 12 indicators on 1,500 unseen BTC 5-min candles. It read them right 99.8% of the time, then scored AUC 0.49. A plain model got 0.52.