List your company

Concept Monitoring on 11 concept monitoring tasks (test)

99.39AccuracyRFM
66.983675.396883.8192.2232Apr 27, 2026

Evaluation results

MethodLinks
RFMAggregation Method=Highest accuracy block, Base Model=Llama-3.1-70B2026.0499.39
RFMAggregation Method=Highest accuracy block, Base Model=Llama-3.1-8B2026.0499.09
RFMAggregation Method=Highest accuracy block, Base Model=Qwen2.5-7B2026.0498.93
RFMAggregation Method=Highest accuracy block, Base Model=Llama-3.2-1B2026.0497.9
LoRAAggregation Method=Highest accuracy block, Base Model=Llama-3.1-70B2026.0497.37
RFMAggregation Method=Average over all blocks, Base Model=Llama-3.1-70B2026.0497.19
ReFTAggregation Method=Highest accuracy block, Base Model=Llama-3.1-70B2026.0496.19
RFMAggregation Method=Average over all blocks, Base Model=Qwen2.5-7B2026.0496.14
RFMAggregation Method=Average over all blocks, Base Model=Llama-3.1-8B2026.0495.13
LoRAAggregation Method=Highest accuracy block, Base Model=Qwen2.5-7B2026.0494.86
ReFTAggregation Method=Highest accuracy block, Base Model=Qwen2.5-7B2026.0494.6
LoRAAggregation Method=Highest accuracy block, Base Model=Llama-3.1-8B2026.0493.26
ReFTAggregation Method=Highest accuracy block, Base Model=Llama-3.1-8B2026.0493.16
RFMAggregation Method=Average over all blocks, Base Model=Llama-3.2-1B2026.0493.02
LoRAAggregation Method=Highest accuracy block, Base Model=Llama-3.2-1B2026.0489.57
ReFTAggregation Method=Highest accuracy block, Base Model=Llama-3.2-1B2026.0488.79
LoRAAggregation Method=Average over all blocks, Base Model=Llama-3.1-70B2026.0474.41
ReFTAggregation Method=Average over all blocks, Base Model=Llama-3.1-70B2026.0473.77
LoRAAggregation Method=Average over all blocks, Base Model=Llama-3.1-8B2026.0471.89
LoRAAggregation Method=Average over all blocks, Base Model=Qwen2.5-7B2026.0470.63
LoRAAggregation Method=Average over all blocks, Base Model=Llama-3.2-1B2026.0469.81
ReFTAggregation Method=Average over all blocks, Base Model=Qwen2.5-7B2026.0468.85
ReFTAggregation Method=Average over all blocks, Base Model=Llama-3.1-8B2026.0468.75
ReFTAggregation Method=Average over all blocks, Base Model=Llama-3.2-1B2026.0468.23