List your company

Question Answering on Qasper (F1)

36.9F1 ScoreLSTM-MAS
-0.73769.033718.80528.5763Jan 17, 2026Jan 18, 2026Jan 19, 2026Jan 21, 2026Jan 22, 2026Jan 23, 2026Jan 25, 2026

Evaluation results

MethodLinks
LSTM-MASBackbone=Phi-3-small-8k-instruct, Context Window=8K2026.0136.9
CoABackbone=Llama-3.1-8B-Instruct, Context Window=8K2026.0135.68
LSTM-MASBackbone=Qwen2.5-7B, Context Window=8K2026.0132.85
LSTM-MASBackbone=Llama-3.1-8B-Instruct, Context Window=8K2026.0132.24
LSTM-MASBackbone=Qwen2.5-32B, Context Window=8K2026.0131.74
LSTM-MASBackbone=Phi-3.5-mini-instruct, Context Window=8K2026.0131.04
CoABackbone=Phi-3-small-8k-instruct, Context Window=8K2026.0130.03
CoABackbone=Phi-3.5-mini-instruct, Context Window=8K2026.0129.89
CoABackbone=Llama-3.2-3B-Instruct, Context Window=8K2026.0128.34
VanillaBackbone=Qwen2.5-14B, Context Window=8K2026.0127.52
VanillaBackbone=Llama-3.1-8B-Instruct, Context Window=8K2026.0124.78
LSTM-MASBackbone=Qwen2.5-1.5B, Context Window=8K2026.0124.17
LSTM-MASBackbone=Qwen2.5-14B, Context Window=8K2026.0123.93
CoABackbone=Qwen2.5-7B, Context Window=8K2026.0122.86
RAG-RerankModel=Mistral-7B-Instruct-v0.32026.0122.51
HyDE-RerankModel=Mistral-7B-Instruct-v0.32026.0122.3
BM25Model=Mistral-7B-Instruct-v0.32026.0121.9
S3-HybridModel=Mistral-7B-Instruct-v0.32026.0121.87
LSTM-MASBackbone=Llama-3.2-3B-Instruct, Context Window=8K2026.0121.81
FullKVModel=Mistral-7B-Instruct-v0.32026.0121.77
RAG-RerankModel=Llama-3.1-8B-Instruct2026.0121.73
HyDE-RerankModel=Llama-3.1-8B-Instruct2026.0121.73
S3-HybridModel=Llama-3.1-8B-Instruct2026.0121.35
BM25Model=Llama-3.1-8B-Instruct2026.0121.34
CoABackbone=Qwen2.5-1.5B, Context Window=8K2026.0121.31
VanillaBackbone=Llama-3.2-3B-Instruct, Context Window=8K2026.0121.23
FullKVModel=Llama-3.1-8B-Instruct2026.0120.9
RAG-RerankModel=Qwen2-7B-Instruct2026.0120.44
S3-PureModel=Mistral-7B-Instruct-v0.32026.0120.42
HyDE-RerankModel=Qwen2-7B-Instruct2026.0120.42
BM25Model=Qwen2-7B-Instruct2026.0120.31
S3-PureModel=Llama-3.1-8B-Instruct2026.0120.08
S3-HybridModel=Qwen2-7B-Instruct2026.0120.04
FullKVModel=Qwen2-7B-Instruct2026.0119.73
VanillaBackbone=Phi-3-small-8k-instruct, Context Window=8K2026.0119
S3-PureModel=Qwen2-7B-Instruct2026.0117.24
CoABackbone=Qwen2.5-32B, Context Window=8K2026.0116.91
CoABackbone=Qwen2.5-14B, Context Window=8K2026.0116.69
LONGAGENTBackbone=Qwen2.5-32B, Context Window=8K2026.0115.08
LONGAGENTBackbone=Qwen2.5-14B, Context Window=8K2026.0112.51
LONGAGENTBackbone=Llama-3.1-8B-Instruct, Context Window=8K2026.0112.46
LightRAGBackbone=Qwen2.5-1.5B2026.0112.44
RAG-DenseModel=Mistral-7B-Instruct-v0.32026.0111.43
LightRAGBackbone=Llama-3.2-3B-Instruct2026.0111.41
LSTM-MASBackbone=Qwen2.5-0.5B, Context Window=8K2026.0111.17
RAG-DenseModel=Qwen2-7B-Instruct2026.0111.15
VanillaBackbone=Qwen2.5-32B, Context Window=8K2026.0110.91
RAG-DenseModel=Llama-3.1-8B-Instruct2026.0110.56
VanillaBackbone=Phi-3.5-mini-instruct, Context Window=8K2026.0110.12
LONGAGENTBackbone=Phi-3.5-mini-instruct, Context Window=8K2026.019.62
VanillaBackbone=Qwen2.5-7B, Context Window=8K2026.019.56
LONGAGENTBackbone=Qwen2.5-1.5B, Context Window=8K2026.019.44
LONGAGENTBackbone=Phi-3-small-8k-instruct, Context Window=8K2026.019.28
CoABackbone=Qwen2.5-0.5B, Context Window=8K2026.019.02
LONGAGENTBackbone=Qwen2.5-7B, Context Window=8K2026.018.09
LightRAGBackbone=Llama-3.1-8B-Instruct2026.017.42
LONGAGENTBackbone=Llama-3.2-3B-Instruct, Context Window=8K2026.014.01
VanillaBackbone=Qwen2.5-1.5B, Context Window=8K2026.013.65
LightRAGBackbone=Qwen2.5-0.5B2026.013.09
VanillaBackbone=Qwen2.5-0.5B, Context Window=8K2026.012.03
LONGAGENTBackbone=Qwen2.5-0.5B, Context Window=8K2026.010.71