List your company

Watermark Detection and Quality Evaluation on Aggregate (MMLU, HellaSwag, ARC-C, GPQA, MBPP, GSM8K)

0.99TPR@1DLM watermark
-0.03960.22770.4950.7623Sep 29, 2025

Evaluation results

MethodLinks
DLM watermarkModel=LLADA-8B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.991.98.4356
DLM watermarkModel=LLADA-8B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.991.88.653.7
DLM watermarkModel=LLADA-8B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.991.88.5954.5
DLM watermarkModel=DREAM-7B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.992.327.7650.1
DLM watermarkModel=DREAM-7B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.992.187.8547
DLM watermarkModel=DREAM-7B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.992.157.947.6
BaselineModel=LLADA-8B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.831.948.3758.5
BaselineModel=DREAM-7B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.742.187.9435.5
BaselineModel=DREAM-7B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.72.238.232.9
BaselineModel=LLADA-8B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.691.868.5155.7
BaselineModel=LLADA-8B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.631.938.4855.5
BaselineModel=DREAM-7B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.090.492.277.9535.8
UnwatermarkedModel=LLADA-8B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.568.9559.4
UnwatermarkedModel=LLADA-8B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.568.9559.4
UnwatermarkedModel=LLADA-8B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.568.9559.4
UnwatermarkedModel=DREAM-7B, Context C={-1}, Strength δ=4, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.948.4550.4
UnwatermarkedModel=DREAM-7B, Context C={-1, 1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.948.4550.4
UnwatermarkedModel=DREAM-7B, Context C={-2, -1}, Strength δ=5, Generation Temperature=0.5, Benchmark Temperature=0.12025.0901.948.4550.4