List your company

Multi-hop Question Answering on HotpotQA (EM, F1, Faithfulness)

75.97F1ACPS
28.70240.973553.24565.5165Jan 13, 2026Jan 14, 2026Jan 15, 2026Jan 16, 2026

Evaluation results

MethodLinks
ACPSModel=GPT-3.5-turbo2026.0175.9761.31—
CPModel=GPT-3.5-turbo2026.0172.5157.05—
ACPSModel=LLaMA-32026.0170.2256.67—
CPModel=LLaMA-32026.0169.9855—
CoT-SCModel=GPT-3.5-turbo2026.0169.0354.74—
DeCoTModel=GPT-3.5-turbo2026.0168.3553.91—
CADModel=GPT-3.5-turbo2026.0167.4552.66—
ACPSModel=Ministral-3B2026.0166.6750—
RT-RAGModels=Qwen2.5-14B2026.0166.2451—
RT-RAGModels=GPT-4o-mini2026.0165.2652.5—
LongRAGModels=GPT-4o-mini2026.0164.7451—
ChainRAG (CxtInt)Models=GPT-4o-mini2026.0164.5950—
DeCoTModel=LLaMA-32026.0164.5254.63—
CADModel=LLaMA-32026.0163.9553.08—
ICLModel=GPT-3.5-turbo2026.0163.7246—
LongRAGModels=Qwen2.5-14B2026.0162.6449.5—
ChainRAG (AnsInt)Models=GPT-4o-mini2026.0160.7346—
NaiveRAG w/ QDModels=GPT-4o-mini2026.016043.5—
IRCoTModels=GPT-4o-mini2026.0159.9245—
SoTModel=GPT-3.5-turbo2026.0159.0440.83—
ItER-RETGENModels=Qwen2.5-14B2026.0158.6344.5—
CoTModel=GPT-3.5-turbo2026.0158.640.54—
NaiveRAGModels=Qwen2.5-14B2026.0157.9643.5—
ItER-RETGENModels=GPT-4o-mini2026.0157.7742—
NaiveRAGModels=GPT-4o-mini2026.0156.9242—
HippoRAG w/ IRCoTModels=GPT-4o-mini2026.0156.1240—
CoT-SCModel=LLaMA-32026.0155.9440.88—
SoTModel=LLaMA-32026.0154.4439.17—
SuReModels=GPT-4o-mini2026.0152.837.5—
NaiveRAG w/ QDModels=Qwen2.5-14B2026.0150.9538.5—
Self-AskModels=GPT-4o-mini2026.0150.6438—
DeCoTModel=Ministral-3B2026.0150.2135.65—
CoTModel=LLaMA-32026.0150.239.16—
CoT-SCModel=Ministral-3B2026.0149.1633.33—
CADModel=Ministral-3B2026.0149.0135.23—
CPModel=Ministral-3B2026.0148.9838.03—
IRCoTModels=Qwen2.5-14B2026.0148.7936.5—
SuReModels=Qwen2.5-14B2026.0148.2133—
CoTModel=Ministral-3B2026.0147.231.58—
HippoRAG w/ IRCoTModels=Qwen2.5-14B2026.0147.1135.5—
SoTModel=Ministral-3B2026.0146.3731.19—
Self-AskModels=Qwen2.5-14B2026.0145.1235—
ICLModel=LLaMA-32026.0143.1832.55—
ICLModel=Ministral-3B2026.0142.0626.92—
CoTModels=GPT-4o-mini2026.0139.7428—
DirectModels=GPT-4o-mini2026.0137.8527.5—
CoTModels=Qwen2.5-14B2026.0132.0321.5—
DirectModels=Qwen2.5-14B2026.0130.5220.5—