List your company

Error Prediction on AmbigQA (val)

69.8PRRSpectrum-Qwen3-14B
-1.33617.13235.654.068Jun 1, 2026

Evaluation results

MethodLinks
Spectrum-Qwen3-14BFeatures=all UQ, Strategy=latent gated2026.0669.8—
Llama-3.1-8BFeatures=all UQ, Strategy=baseline2026.0669.6—
Spectrum-Qwen3-14BFeatures=all UQ, Strategy=baseline2026.0669.6—
Llama-3.1-8BFeatures=all UQ, Strategy=latent gated2026.0669.5—
Spectrum-Qwen3-14BFeatures=all UQ, Strategy=latent selective2026.0668.8—
Llama-3.1-8BFeatures=SAR, Strategy=latent gated2026.0668.5—
Llama-3.1-8BFeatures=SAR, Strategy=baseline2026.0668.3—
Llama-3.1-8BFeatures=all UQ, Strategy=latent selective2026.0668.1—
Llama-3.1-8BFeatures=CoCoA, Strategy=latent gated2026.0665.1—
oracle gatedModel=Llama-3.1-8B, Features=all UQ2026.0665.1—
Llama-3.1-8BFeatures=CoCoA, Strategy=baseline2026.0664.4—
latent gatedModel=Llama-3.1-8B, Features=all UQ2026.0663.3—
baselineModel=Llama-3.1-8B, Features=all UQ2026.0663.2—
oracle selectiveModel=Llama-3.1-8B, Features=all UQ2026.0662.7—
latent selectiveModel=Llama-3.1-8B, Features=all UQ2026.0662—
A2Search-7BFeatures=all UQ, Strategy=latent selective2026.0661.6—
A2Search-7BFeatures=all UQ, Strategy=latent gated2026.0661.4—
Spectrum-Llama-3.1-8BFeatures=all UQ, Strategy=latent selective2026.0660.7—
Spectrum-Llama-3.1-8BFeatures=all UQ, Strategy=latent gated2026.0660.5—
Spectrum-Llama-3.1-8BFeatures=all UQ, Strategy=baseline2026.0660.2—
A2Search-7BFeatures=all UQ, Strategy=baseline2026.0659.3—
Qwen3-14BFeatures=all UQ, Strategy=latent gated2026.0658.4—
latent gatedModel=A^2Search-7B, Features=all UQ2026.0658—
latent selectiveModel=A^2Search-7B, Features=all UQ2026.0658—
Qwen3-14BFeatures=all UQ, Strategy=baseline2026.0657.9—
oracle gatedModel=Qwen3-14B, Features=all UQ2026.0657.5—
oracle gatedFeatures=all UQ, Backbone=Qwen3-14B, Supervision=AlignScore2026.0657.5—
oracle gatedModel=A^2Search-7B, Features=all UQ2026.0657.3—
latent gatedModel=Qwen3-14B, Features=all UQ2026.0657.1—
latent gatedFeatures=all UQ, Backbone=Qwen3-14B, Supervision=AlignScore2026.0657.1—
oracle gatedModel=Spectrum-Qwen3-14B, Features=all UQ2026.0656.5—
Qwen3-14BFeatures=all UQ, Strategy=latent selective2026.0656.4—
latent selectiveModel=Qwen3-14B, Features=all UQ2026.0656.2—
latent selectiveFeatures=all UQ, Backbone=Qwen3-14B, Supervision=AlignScore2026.0656.2—
baselineModel=Spectrum-Qwen3-14B, Features=all UQ2026.0656—
latent gatedModel=Spectrum-Qwen3-14B, Features=all UQ2026.0656—
latent selectiveModel=Spectrum-Qwen3-14B, Features=all UQ2026.0655.6—
baselineModel=Qwen3-14B, Features=all UQ2026.0655.5—
baselineFeatures=all UQ, Backbone=Qwen3-14B, Supervision=AlignScore2026.0655.5—
latent gatedFeatures=CoCoA, Backbone=Qwen3-14B, Supervision=AlignScore2026.0655.3—
oracle gatedFeatures=CoCoA, Backbone=Qwen3-14B, Supervision=AlignScore2026.0655.3—
oracle selectiveModel=A^2Search-7B, Features=all UQ2026.0654.5—
baselineModel=A^2Search-7B, Features=all UQ2026.0654.3—
oracle gatedFeatures=SAR, Backbone=Qwen3-14B, Supervision=AlignScore2026.0654.3—
latent gatedFeatures=SAR, Backbone=Qwen3-14B, Supervision=AlignScore2026.0653.9—
baselineFeatures=CoCoA, Backbone=Qwen3-14B, Supervision=AlignScore2026.0653.8—
oracle selectiveModel=Qwen3-14B, Features=all UQ2026.0653.7—
oracle selectiveFeatures=all UQ, Backbone=Qwen3-14B, Supervision=AlignScore2026.0653.7—
oracle selectiveModel=Spectrum-Qwen3-14B, Features=all UQ2026.0653.6—
oracle selectiveFeatures=SAR, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652.8—
oracle selectiveFeatures=CoCoA, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652.6—
latent selectiveFeatures=SAR, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652.6—
baselineFeatures=SAR, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652.5—
latent selectiveFeatures=CoCoA, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652.4—
latent gatedFeatures=Sem. energy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652—
latent selectiveFeatures=Sem. energy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0652—
latent gatedFeatures=Sem. entropy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0651.3—
latent gatedFeatures=MSP, Backbone=Qwen3-14B, Supervision=AlignScore2026.0651.2—
latent selectiveFeatures=Sem. entropy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0650.8—
oracle gatedFeatures=MSP, Backbone=Qwen3-14B, Supervision=AlignScore2026.0650.4—
latent selectiveFeatures=MSP, Backbone=Qwen3-14B, Supervision=AlignScore2026.0649.2—
oracle gatedModel=Spectrum-Llama-3.1-8B, Features=all UQ2026.0648.3—
baselineFeatures=MSP, Backbone=Qwen3-14B, Supervision=AlignScore2026.0648.3—
oracle selectiveFeatures=MSP, Backbone=Qwen3-14B, Supervision=AlignScore2026.0647.8—
oracle gatedFeatures=Sem. entropy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0647.1—
latent gatedModel=Spectrum-Llama-3.1-8B, Features=all UQ2026.0646.2—
baselineModel=Spectrum-Llama-3.1-8B, Features=all UQ2026.0646.1—
oracle gatedFeatures=Sem. energy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0646—
latent selectiveModel=Spectrum-Llama-3.1-8B, Features=all UQ2026.0645.7—
oracle selectiveModel=Spectrum-Llama-3.1-8B, Features=all UQ2026.0645—
oracle selectiveFeatures=Sem. entropy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0644.6—
oracle selectiveFeatures=Sem. energy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0643.2—
baselineFeatures=Sem. entropy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0642.1—
baselineFeatures=Sem. energy, Backbone=Qwen3-14B, Supervision=AlignScore2026.0639.2—
oracle gatedModel=Llama-3.1-8B, Features=MI2026.0619.9—
baselineModel=Llama-3.1-8B, Features=MI2026.0618.9—
latent gatedModel=Llama-3.1-8B, Features=MI2026.0618.9—
oracle selectiveModel=Llama-3.1-8B, Features=MI2026.0618.2—
latent selectiveModel=Llama-3.1-8B, Features=MI2026.0618.1—
latent gatedModel=Qwen3-14B, Features=MI2026.067.8—
latent selectiveModel=Qwen3-14B, Features=MI2026.067.8—
latent gatedFeatures=MI, Backbone=Qwen3-14B, Supervision=AlignScore2026.067.8—
latent selectiveFeatures=MI, Backbone=Qwen3-14B, Supervision=AlignScore2026.067.8—
oracle gatedModel=Qwen3-14B, Features=MI2026.065.8—
oracle selectiveModel=Qwen3-14B, Features=MI2026.065.8—
oracle gatedFeatures=MI, Backbone=Qwen3-14B, Supervision=AlignScore2026.065.8—
oracle selectiveFeatures=MI, Backbone=Qwen3-14B, Supervision=AlignScore2026.065.8—
Qwen3-14BFeatures=MI, Strategy=latent gated2026.063—
baselineModel=Qwen3-14B, Features=MI2026.061.4—
baselineFeatures=MI, Backbone=Qwen3-14B, Supervision=AlignScore2026.061.4—
A^2Search-7B + all UQmodel=A^2Search-7B, features=all UQ, setting=baseline2026.06—75.7
A^2Search-7B + all UQmodel=A^2Search-7B, features=all UQ, setting=latent gated2026.06—76.7
A^2Search-7B + MImodel=A^2Search-7B, features=MI, setting=latent gated2026.06—55.5
A^2Search-7B + MImodel=A^2Search-7B, features=MI, setting=latent selective2026.06—55.5
Llama-3.1-8B + all UQmodel=Llama-3.1-8B, features=all UQ, setting=baseline2026.06—79.9
Llama-3.1-8B + all UQmodel=Llama-3.1-8B, features=all UQ, setting=latent gated2026.06—79.8
Llama-3.1-8B + all UQmodel=Llama-3.1-8B, features=all UQ, setting=oracle gated2026.06—79.9
Llama-3.1-8B + all UQmodel=Llama-3.1-8B, features=all UQ, setting=latent selective2026.06—79.7
Llama-3.1-8B + all UQmodel=Llama-3.1-8B, features=all UQ, setting=oracle selective2026.06—79.7
Llama-3.1-8B + CoCoAmodel=Llama-3.1-8B, features=CoCoA, setting=baseline2026.06—76.8