New graph-neural-network flavor tagger for Belle II and measurement of sin <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mn>2</mml:mn><mml:msub><mml:mi>ϕ</mml:mi><mml:mn>1</mml:mn></mml:msub></mml:math> in <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:msup><mml:mi>B</mml:mi><mml:mn>0</mml:mn></mml:msup><mml:mo stretchy="false">→</mml:mo><mml:mi>J</mml:mi><mml:mo>/</mml:mo><mml:mi>ψ</mml:mi><mml:msubsup><mml:mi>K</mml:mi><mml:mi mathvariant="normal">S</mml:mi><mml:mn>0</mml:mn></mml:msubsup></mml:math> decays

We present GFlaT, a new algorithm that uses a graph-neural-network to determine the flavor of neutral <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" display="inline"><a:mi>B</a:mi></a:math> mesons produced in <c:math xmlns:c="http://www.w3.org/1998/Math/MathML" display="inline"><c:mi mathvariant="normal">ϒ</c:mi><c:mo stretchy="false">(</c:mo><c:mn>4</c:mn><c:mi>S</c:mi><c:mo stretchy="false">)</c:mo></c:math> decays. It improves previous algorithms by using the information from all charged final-state particles and the relations between them. We evaluate its performance using <h:math xmlns:h="http://www.w3.org/1998/Math/MathML" display="inline"><h:mi>B</h:mi></h:math> decays to flavor-specific hadronic final states reconstructed in a <j:math xmlns:j="http://www.w3.org/1998/Math/MathML" display="inline"><j:mn>362</j:mn><j:mtext> </j:mtext><j:mtext> </j:mtext><j:msup><j:mi>fb</j:mi><j:mrow><j:mo>−</j:mo><j:mn>1</j:mn></j:mrow></j:msup></j:math> sample of electron-positron collisions collected at the <l:math xmlns:l="http://www.w3.org/1998/Math/MathML" display="inline"><l:mi mathvariant="normal">ϒ</l:mi><l:mo stretchy="false">(</l:mo><l:mn>4</l:mn><l:mi>S</l:mi><l:mo stretchy="false">)</l:mo></l:math> resonance with the Belle II detector at the SuperKEKB collider. We achieve an effective tagging efficiency of <q:math xmlns:q="http://www.w3.org/1998/Math/MathML" display="inline"><q:mo stretchy="false">(</q:mo><q:mn>37.40</q:mn><q:mo>±</q:mo><q:mn>0.43</q:mn><q:mo>±</q:mo><q:mn>0.36</q:mn><q:mo>%</q:mo><q:mo stretchy="false">)</q:mo></q:math>, where the first uncertainty is statistical and the second systematic, which is 18% better than the previous Belle II algorithm. Demonstrating the algorithm, we use <u:math xmlns:u="http://www.w3.org/1998/Math/MathML" display="inline"><u:msup><u:mi>B</u:mi><u:mn>0</u:mn></u:msup><u:mo stretchy="false">→</u:mo><u:mrow><u:mi>J</u:mi><u:mo>/</u:mo><u:mi>ψ</u:mi></u:mrow><u:msubsup><u:mi>K</u:mi><u:mi mathvariant="normal">S</u:mi><u:mn>0</u:mn></u:msubsup></u:math> decays to measure the mixing-induced and direct <y:math xmlns:y="http://www.w3.org/1998/Math/MathML" display="inline"><y:mi>C</y:mi><y:mi>P</y:mi></y:math> violation parameters, <ab:math xmlns:ab="http://www.w3.org/1998/Math/MathML" display="inline"><ab:mi>S</ab:mi><ab:mo>=</ab:mo><ab:mo stretchy="false">(</ab:mo><ab:mn>0.724</ab:mn><ab:mo>±</ab:mo><ab:mn>0.035</ab:mn><ab:mo>±</ab:mo><ab:mn>0.009</ab:mn><ab:mo stretchy="false">)</ab:mo></ab:math> and <eb:math xmlns:eb="http://www.w3.org/1998/Math/MathML" display="inline"><eb:mi>C</eb:mi><eb:mo>=</eb:mo><eb:mo stretchy="false">(</eb:mo><eb:mo>−</eb:mo><eb:mn>0.035</eb:mn><eb:mo>±</eb:mo><eb:mn>0.026</eb:mn><eb:mo>±</eb:mo><eb:mn>0.029</eb:mn><eb:mo stretchy="false">)</eb:mo></eb:math>. Published by the American Physical Society 2024