List your company

Science Question Answering on GPQA (Context Length Analysis)

18,851.29Avg Context Length (tokens)PRD
1,916.56446,313.079710,709.59515,106.1103Jul 5, 2025

Evaluation results

MethodLinks
PRDType=Full Debate2025.0718,851.29
PRDType=Full Debate2025.0718,837.51
RECONCILEType=Full Debate2025.0714,107.64
RECONCILEType=Full Debate2025.0714,061.76
MLDType=Full Debate2025.0712,947.8
MLDType=Full Debate2025.0712,868.37
ChatEvalType=Full Debate2025.0712,369.62
ChatEvalType=Full Debate2025.0712,252.56
GDType=Part Debate2025.078,906.51
GDType=Part Debate2025.078,817.46
NDType=Part Debate2025.078,016.54
NDType=Part Debate2025.077,971.6
OursType=Part Debate2025.076,503.76
OursType=Part Debate2025.076,340.26
MaVType=No Debate2025.072,653.92
MaVType=No Debate2025.072,567.9