List your company

Question Answering on MedQA (F1, avg_cost)

72F1 Score (%)Top-k
43.9251.2158.565.79Jun 1, 2026

Evaluation results

MethodLinks
Top-kBackbone=Qwen3-8B, Bgt/k=102026.0672—
RedundancyBackbone=Qwen3-8B, Bgt/k=302026.06711.8
Top-kBgt/k=10, Backbone=Llama3.1-8B-instruct2026.0662—
Top-kBgt/k=30, Backbone=Llama3.1-8B-instruct2026.0659—
RedundancyBgt/k=10, Backbone=Llama3.1-8B-instruct2026.06591.68
RedundancyBgt/k=25, Backbone=Llama3.1-8B-instruct2026.06591.76
KnapsackBgt/k=15, Backbone=Llama3.1-8B-instruct2026.065814.3
MMRBgt/k=25, Backbone=Llama3.1-8B-instruct2026.06589.34
MMRBgt/k=∞, Backbone=Llama3.1-8B-instruct2026.06589.37
RelevanceBgt/k=5, Backbone=Llama3.1-8B-instruct2026.06575
GreedyBgt/k=15, Backbone=Llama3.1-8B-instruct2026.065713.73
KnapsackBgt/k=25, Backbone=Llama3.1-8B-instruct2026.065724.31
RedundancyBgt/k=15, Backbone=Llama3.1-8B-instruct2026.06571.76
Top-kBgt/k=5, Backbone=Llama3.1-8B-instruct2026.0656—
Top-kBgt/k=15, Backbone=Llama3.1-8B-instruct2026.0656—
RelevanceBgt/k=0, Backbone=Llama3.1-8B-instruct2026.06560
RelevanceBgt/k=15, Backbone=Llama3.1-8B-instruct2026.065614.9
MMRBgt/k=10, Backbone=Llama3.1-8B-instruct2026.06567.58
Vanilla LLMBackbone=Llama3.1-8B-instruct2026.0655—
RelevanceBgt/k=∞, Backbone=Llama3.1-8B-instruct2026.065561.7
GreedyBgt/k=30, Backbone=Llama3.1-8B-instruct2026.065528.57
Top-kBgt/k=25, Backbone=Llama3.1-8B-instruct2026.0654—
GreedyBgt/k=20, Backbone=Llama3.1-8B-instruct2026.065418.54
KnapsackBgt/k=5, Backbone=Llama3.1-8B-instruct2026.06544.87
KnapsackBgt/k=∞, Backbone=Llama3.1-8B-instruct2026.065461.73
RedundancyBgt/k=20, Backbone=Llama3.1-8B-instruct2026.06541.76
RelevanceBgt/k=10, Backbone=Llama3.1-8B-instruct2026.065310
RelevanceBgt/k=30, Backbone=Llama3.1-8B-instruct2026.065329.8
GreedyBgt/k=0, Backbone=Llama3.1-8B-instruct2026.06530
KnapsackBgt/k=10, Backbone=Llama3.1-8B-instruct2026.06539.58
MMRBgt/k=15, Backbone=Llama3.1-8B-instruct2026.06538.81
MMRBgt/k=20, Backbone=Llama3.1-8B-instruct2026.06539.18
MMRBgt/k=30, Backbone=Llama3.1-8B-instruct2026.06539.37
Top-kBgt/k=20, Backbone=Llama3.1-8B-instruct2026.0652—
RelevanceBgt/k=25, Backbone=Llama3.1-8B-instruct2026.065225
KnapsackBgt/k=30, Backbone=Llama3.1-8B-instruct2026.065229.26
MMRBgt/k=5, Backbone=Llama3.1-8B-instruct2026.06524.66
RedundancyBgt/k=5, Backbone=Llama3.1-8B-instruct2026.06521.18
GreedyBgt/k=5, Backbone=Llama3.1-8B-instruct2026.06514.78
KnapsackBgt/k=0, Backbone=Llama3.1-8B-instruct2026.06510
KnapsackBgt/k=20, Backbone=Llama3.1-8B-instruct2026.065119.2
RedundancyBgt/k=30, Backbone=Llama3.1-8B-instruct2026.06511.76
RedundancyBgt/k=∞, Backbone=Llama3.1-8B-instruct2026.06511.76
GreedyBgt/k=25, Backbone=Llama3.1-8B-instruct2026.065023.5
GreedyBgt/k=∞, Backbone=Llama3.1-8B-instruct2026.065061.73
GreedyBgt/k=10, Backbone=Llama3.1-8B-instruct2026.06489.22
RelevanceBgt/k=20, Backbone=Llama3.1-8B-instruct2026.064719.9
MMRBgt/k=0, Backbone=Llama3.1-8B-instruct2026.06470
RedundancyBgt/k=0, Backbone=Llama3.1-8B-instruct2026.06470
Vanilla LLMBackbone=Qwen3-8B, Bgt/k=-2026.0645—