Table 9 Comparison of discriminator strategies on the BulkModulus dataset

From: SLM-MATRIX: a multi-agent trajectory reasoning and verification framework for enhancing language models in materials data extraction

Discriminator Strategy

Accuracy(%)

Manual Adjudication Rate (%)

Mask–Reconstruct Discriminator

92.85% (±2.05%)

2.98% (±0.85%)

Direct Path Evaluation

88.45% (±1.80%)

5.77% (±1.71%)

Back to article page