List your company

Semantic Uncertainty Stability Prediction on BBEH

97DCS AUROCOutput Embedding
43.9657.7371.585.27May 7, 2026

Evaluation results

MethodLinks
Output EmbeddingLLM=Sky-OR1 32B2026.0597
Output EmbeddingLLM=Llama 8B2026.0596
Output EmbeddingLLM=DS-R1 32B2026.0596
Last Input TokenLLM=Sky-OR1 32B2026.0596
Last Input TokenLLM=DS-R1 32B2026.0595
Soft ThinkingLLM=DS-R1 32B2026.0595
Soft ThinkingLLM=Sky-OR1 32B2026.0595
Last Input TokenLLM=Llama 8B2026.0594
Soft ThinkingLLM=Llama 8B2026.0594
Input EmbeddingLLM=DS-R1 32B2026.0594
Input EmbeddingLLM=Llama 8B2026.0593
Input EmbeddingLLM=Sky-OR1 32B2026.0593
Latent ThinkingLLM=Llama 8B2026.0592
Input EmbeddingLLM=Llama 70B2026.0592
Latent ThinkingLLM=DS-R1 32B2026.0592
Latent ThinkingLLM=Sky-OR1 32B2026.0592
Soft Thinking with NoiseLLM=Llama 8B2026.0590
Output EmbeddingLLM=Llama 70B2026.0589
Last Input TokenLLM=Llama 70B2026.0589
Latent ThinkingLLM=Llama 70B2026.0587
Soft Thinking with NoiseLLM=Sky-OR1 32B2026.0586
Soft ThinkingLLM=Llama 70B2026.0585
Soft Thinking with NoiseLLM=DS-R1 32B2026.0585
Soft Thinking with NoiseLLM=Llama 70B2026.0584
Output EmbeddingLLM=GPT-OSS 20B2026.0574
Latent ThinkingLLM=GPT-OSS 20B2026.0559
Input EmbeddingLLM=GPT-OSS 20B2026.0559
Last Input TokenLLM=GPT-OSS 20B2026.0558
Random VectorLLM=GPT-OSS 20B2026.0556
Soft ThinkingLLM=GPT-OSS 20B2026.0555
Random VectorLLM=Llama 8B2026.0552
Random VectorLLM=Llama 70B2026.0550
Random VectorLLM=DS-R1 32B2026.0550
Random VectorLLM=Sky-OR1 32B2026.0549
Soft Thinking with NoiseLLM=GPT-OSS 20B2026.0546