benwtrent commented on issue #13350: URL: https://github.com/apache/lucene/issues/13350#issuecomment-2137366217
@jmazanec15 this is accounted for in the corrections. The moving from signed to unsigned is still just a linear transformation, we are not manually flipping signs, but instead doing a full linear scale. Assuming your vectors are in order that you provided `[[-6][-2][3][6]]` Their calculated dot-product corrections are `[18.0, -5.8750076, -35.787956, -54.0]` For query vector `[-5]` its correction is `11.95908` The overall quantile multiplier is `0.008928018` To calculate the corrected dot-product score `quantizedDotProduct * multiplier + queryCorrection + vectorCorrection` The raw dot-products (in order of the vectors) ``` [30.0, 10.0, -15.0, -30.0] ``` The quantized dot-products with corrections ``` [29.95908, 10.208817, -14.499098, -29.568478] ``` Of course, the quantized scores without corrections (thus not accounting for linear shift and breaking max-inner product score scaling) ``` [0.0, 462.0, 1045.0, 1397.0] ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
