TCSP2.0
A previous version of the template-based crystal structure prediction system that uses sequential text substitution for element mapping and exhibits systematic deficiencies such as neglect of stoichiometric coefficients and symbol-level cascading errors.
Timeline (last 90 days)
Of the 2347 structures selected by the T180 benchmark, 469 (20%) had incorrect compositions in TCSP2.0.
The root cause of TCSP2.0's composition errors lies in element mapping that ignores stoichiometric coefficients and in sequential text replacement that causes symbol cascading or corruption.
On the T180 benchmark, TCSP3.0 achieves Top-1 and Top-20 match rates of 72.2% and 90.6%, respectively, whereas TCSP2.0 reaches only 28.3% and 45.6%.
All Facts (3)
Of the 2347 structures selected by the T180 benchmark, 469 (20%) had incorrect compositions in TCSP2.0.
50%UnverifiedThe root cause of TCSP2.0's composition errors lies in element mapping that ignores stoichiometric coefficients and in sequential text replacement that causes symbol cascading or corruption.
50%UnverifiedOn the T180 benchmark, TCSP3.0 achieves Top-1 and Top-20 match rates of 72.2% and 90.6%, respectively, whereas TCSP2.0 reaches only 28.3% and 45.6%.
50%