LONGXIA

? · · lstm_s3

token id: #185
symbol: LONGXIA
variant: –
chain
address
cohort: datapoint (mid-window listing)
first bar: 2026-08-27
predictions: 144

Track record at +1d

assessed

47

of 48 predictions shown

direction hit

57.4%

alpha space; 50% is a coin flip

MAE (alpha)

0.2309

mean absolute log-return error

mean pred / actual

-0.0131

actual -0.0320

History

as of	split	anchor px	pred px	pred Δ	actual px	actual Δ	alpha err	hit	β90d
2026-10-22	oos	0.02106	0.020967	-0.44%	–	pending	–	–	1.35
2026-10-21	oos	0.021215	0.021127	-0.41%	0.02106	-0.73%	0.0063	✓	1.35
2026-10-20	oos	0.021195	0.021104	-0.43%	0.021215	+0.09%	0.0124	✓	1.35
2026-10-19	oos	0.020355	0.020247	-0.53%	0.021195	+4.13%	0.0280	✗	1.33
2026-10-18	oos	0.02088	0.020757	-0.59%	0.020355	-2.51%	0.0168	✓	1.33
2026-10-17	oos	0.02746	0.027353	-0.39%	0.02088	-23.96%	0.2811	✓	1.44
2026-10-16	oos	0.029085	0.028976	-0.37%	0.02746	-5.59%	0.0517	✓	1.44
2026-10-15	oos	0.029995	0.029887	-0.36%	0.029085	-3.03%	0.0163	✓	1.43
2026-10-14	oos	0.029625	0.029504	-0.41%	0.029995	+1.25%	0.0068	✗	1.43
2026-10-13	oos	0.04862	0.048321	-0.61%	0.029625	-39.07%	0.4721	✓	0.92
2026-10-12	oos	0.040295	0.040128	-0.41%	0.04862	+20.66%	0.1793	✗	0.77
2026-10-11	oos	0.04401	0.043808	-0.46%	0.040295	-8.44%	0.0681	✓	0.66
2026-10-10	oos	0.04627	0.046057	-0.46%	0.04401	-4.88%	0.0519	✓	0.69
2026-10-09	oos	0.046	0.045798	-0.44%	0.04627	+0.59%	0.0072	✗	0.69
2026-10-08	oos	0.05196	0.051693	-0.51%	0.046	-11.47%	0.1247	✓	0.77
2026-10-07	oos	0.04116	0.040899	-0.63%	0.05196	+26.24%	0.2598	✗	1.07
2026-10-06	oos	0.06616	0.065639	-0.79%	0.04116	-37.79%	0.4601	✓	0.25
2026-10-05	oos	0.04544	0.045063	-0.83%	0.06616	+45.60%	0.3849	✗	0.34
2026-10-04	oos	0.04529	0.044867	-0.93%	0.04544	+0.33%	0.0157	✗	0.34
2026-10-03	oos	0.049945	0.049471	-0.95%	0.04529	-9.32%	0.0983	✓	0.48
2026-10-02	oos	0.03919	0.038653	-1.37%	0.049945	+27.44%	0.2550	✗	0.46
2026-10-01	oos	0.110175	0.10831	-1.69%	0.03919	-64.43%	1.0163	✓	0.07
2026-09-30	oos	0.0359	0.035358	-1.51%	0.110175	+206.89%	1.1512	✗	-0.96
2026-09-29	oos	0.068975	0.068267	-1.03%	0.0359	-47.95%	0.6437	✓	-1.06
2026-09-28	oos	0.070225	0.069453	-1.10%	0.068975	-1.78%	0.0046	✓	-1.05
2026-09-27	oos	0.08775	0.086874	-1.00%	0.070225	-19.97%	0.2283	✓	-1.29
2026-09-26	oos	0.1038	0.102712	-1.05%	0.08775	-15.46%	0.1567	✓	-1.31
2026-09-25	oos	0.130175	0.128453	-1.32%	0.1038	-20.26%	0.2169	✓	1.00
2026-09-24	oos	0.10344	0.10208	-1.32%	0.130175	+25.85%	0.2466	✗	1.00
2026-09-23	oos	0.16812	0.165059	-1.82%	0.10344	-38.47%	0.4674	✓	1.00
2026-09-22	oos	0.139465	0.137192	-1.63%	0.16812	+20.55%	0.2246	✗	1.00
2026-09-21	oos	0.234285	0.228991	-2.26%	0.139465	-40.47%	0.4913	✓	1.00
2026-09-20	oos	0.238305	0.231781	-2.74%	0.234285	-1.69%	0.0541	✓	1.00
2026-09-19	oos	0.226335	0.219985	-2.81%	0.238305	+5.29%	0.0809	✗	1.00
2026-09-18	oos	0.22224	0.215627	-2.98%	0.226335	+1.84%	0.0440	✗	1.00
2026-09-17	oos	0.23069	0.222571	-3.52%	0.22224	-3.66%	0.0591	✓	1.00
2026-09-16	oos	0.21309	0.205679	-3.48%	0.23069	+8.26%	0.1121	✗	1.00
2026-09-15	oos	0.195995	0.189211	-3.46%	0.21309	+8.72%	0.1115	✗	1.00
2026-09-14	oos	0.16108	0.156222	-3.02%	0.195995	+21.68%	0.2605	✗	1.00
2026-09-13	oos	0.14472	0.140197	-3.13%	0.16108	+11.30%	0.1211	✗	1.00
2026-09-12	oos	0.154585	0.148864	-3.70%	0.14472	-6.38%	0.0222	✓	1.00
2026-09-11	oos	0.07052	0.06983	-0.98%	0.154585	+119.21%	0.7940	✗	1.00
2026-09-10	oos	0.03809	0.037944	-0.38%	0.07052	+85.14%	0.6110	✗	1.00
2026-09-09	oos	0.049155	0.048917	-0.48%	0.03809	-22.51%	0.2276	✓	1.00
2026-09-08	oos	0.05884	0.058488	-0.60%	0.049155	-16.46%	0.1718	✓	1.00
2026-09-07	oos	0.05455	0.054271	-0.51%	0.05884	+7.86%	0.0890	✗	1.00
2026-09-06	oos	0.0741	0.073611	-0.66%	0.05455	-26.38%	0.2840	✓	1.00
2026-09-05	oos	0.08774	0.086864	-1.00%	0.0741	-15.55%	0.1653	✓	1.00

predicted and realised values are beta-adjusted unless labelled otherwise: alpha = r_token − β_90d · r_BTC, so every figure means “performance versus this token’s own BTC beta”, not market direction. Prices are reconstructed under the flat-BTC assumption P̂ = P_last exp(clip(α̂, ±3)). In-sample (train) rows are excluded by default because the model trained on them. Read-only view of predictions.db.