Deceptive Commitment Prediction on Leave-one-environment-out transfer (held-out fifth)
0.755AUROCAttention + PCA final
Evaluation Results
| Method | Links | |
|---|---|---|
| Attention + PCA finalModel=R1-Distill Qwen-7B, Feature Category=Combined2026.05 | 0.755 | |
| Attention + PCA final - mean(prev 4)Model=R1-Distill Qwen-7B, Feature Category=Combined2026.05 | 0.749 | |
| Attention + PCA final - prevModel=R1-Distill Qwen-7B, Feature Category=Combined2026.05 | 0.745 | |
| All attentionModel=R1-Distill Qwen-7B, Feature Category=Attention2026.05 | 0.729 | |
| Attention + PCA finalModel=R1-Distill Qwen-14B, Feature Category=Combined2026.05 | 0.728 | |
| Grounding onlyModel=R1-Distill Qwen-7B, Feature Category=Attention2026.05 | 0.721 | |
| Grounding transition onlyModel=R1-Distill Qwen-7B, Feature Category=Attention2026.05 | 0.72 | |
| Grounding transition onlyModel=R1-Distill Qwen-14B, Feature Category=Attention2026.05 | 0.715 | |
| Raw ActivationModel=R1-Distill Qwen-7B, Feature Category=Activation2026.05 | 0.714 | |
| Attention + PCA final - mean(prev 4)Model=R1-Distill Qwen-14B, Feature Category=Combined2026.05 | 0.713 | |
| Attention + PCA final - prevModel=R1-Distill Qwen-14B, Feature Category=Combined2026.05 | 0.711 | |
| Attention + PCA finalModel=GPT-OSS-20B, Feature Category=Combined2026.05 | 0.708 | |
| All attentionModel=R1-Distill Qwen-14B, Feature Category=Attention2026.05 | 0.707 | |
| Raw ActivationModel=R1-Distill Llama-8B, Feature Category=Activation2026.05 | 0.705 | |
| Attention + PCA finalModel=R1-Distill Llama-8B, Feature Category=Combined2026.05 | 0.701 | |
| PCA finalModel=R1-Distill Qwen-7B, Feature Category=Activation2026.05 | 0.695 | |
| PCA final - prevModel=R1-Distill Qwen-7B, Feature Category=Activation2026.05 | 0.695 | |
| Grounding onlyModel=R1-Distill Qwen-14B, Feature Category=Attention2026.05 | 0.693 | |
| PCA final - mean(prev 4)Model=R1-Distill Qwen-7B, Feature Category=Activation2026.05 | 0.692 | |
| Attention + PCA final - mean(prev 4)Model=R1-Distill Llama-8B, Feature Category=Combined2026.05 | 0.692 | |
| PCA finalModel=R1-Distill Llama-8B, Feature Category=Activation2026.05 | 0.683 | |
| Attention + PCA final - mean(prev 4)Model=GPT-OSS-20B, Feature Category=Combined2026.05 | 0.681 | |
| Concentration transition onlyModel=R1-Distill Qwen-14B, Feature Category=Attention2026.05 | 0.678 | |
| Attention + PCA final - prevModel=R1-Distill Llama-8B, Feature Category=Combined2026.05 | 0.677 | |
| PCA final - mean(prev 4)Model=R1-Distill Llama-8B, Feature Category=Activation2026.05 | 0.675 | |
| Attention + PCA final - prevModel=GPT-OSS-20B, Feature Category=Combined2026.05 | 0.675 | |
| Grounding transition onlyModel=GPT-OSS-20B, Feature Category=Attention2026.05 | 0.674 | |
| All attentionModel=GPT-OSS-20B, Feature Category=Attention2026.05 | 0.673 | |
| Grounding onlyModel=GPT-OSS-20B, Feature Category=Attention2026.05 | 0.672 | |
| Concentration transition onlyModel=R1-Distill Qwen-7B, Feature Category=Attention2026.05 | 0.666 | |
| PCA finalModel=R1-Distill Qwen-14B, Feature Category=Activation2026.05 | 0.657 | |
| Raw ActivationModel=R1-Distill Qwen-14B, Feature Category=Activation2026.05 | 0.653 | |
| PCA final - prevModel=R1-Distill Llama-8B, Feature Category=Activation2026.05 | 0.648 | |
| Grounding onlyModel=R1-Distill Llama-8B, Feature Category=Attention2026.05 | 0.648 | |
| Raw ActivationModel=GPT-OSS-20B, Feature Category=Activation2026.05 | 0.645 | |
| Concentration onlyModel=R1-Distill Qwen-7B, Feature Category=Attention2026.05 | 0.642 | |
| All attentionModel=R1-Distill Llama-8B, Feature Category=Attention2026.05 | 0.641 | |
| Grounding transition onlyModel=R1-Distill Llama-8B, Feature Category=Attention2026.05 | 0.638 | |
| PCA final - mean(prev 4)Model=R1-Distill Qwen-14B, Feature Category=Activation2026.05 | 0.633 | |
| PCA finalModel=GPT-OSS-20B, Feature Category=Activation2026.05 | 0.628 | |
| Concentration onlyModel=GPT-OSS-20B, Feature Category=Attention2026.05 | 0.628 | |
| Concentration onlyModel=R1-Distill Qwen-14B, Feature Category=Attention2026.05 | 0.625 | |
| Concentration transition onlyModel=GPT-OSS-20B, Feature Category=Attention2026.05 | 0.625 | |
| PCA final - prevModel=R1-Distill Qwen-14B, Feature Category=Activation2026.05 | 0.617 | |
| Concentration transition onlyModel=R1-Distill Llama-8B, Feature Category=Attention2026.05 | 0.61 | |
| PCA final - mean(prev 4)Model=GPT-OSS-20B, Feature Category=Activation2026.05 | 0.597 | |
| PCA final - prevModel=GPT-OSS-20B, Feature Category=Activation2026.05 | 0.591 | |
| Concentration onlyModel=R1-Distill Llama-8B, Feature Category=Attention2026.05 | 0.573 | |
| TF-IDF last sentenceModel=R1-Distill Qwen-7B, Feature Category=Lexical Baselines2026.05 | 0.544 | |
| TF-IDF last sentenceModel=R1-Distill Qwen-14B, Feature Category=Lexical Baselines2026.05 | 0.542 | |
| TF-IDF prefixModel=R1-Distill Qwen-7B, Feature Category=Lexical Baselines2026.05 | 0.53 | |
| TF-IDF last sentenceModel=R1-Distill Llama-8B, Feature Category=Lexical Baselines2026.05 | 0.491 | |
| TF-IDF prefixModel=R1-Distill Qwen-14B, Feature Category=Lexical Baselines2026.05 | 0.49 | |
| TF-IDF last sentenceModel=GPT-OSS-20B, Feature Category=Lexical Baselines2026.05 | 0.482 | |
| TF-IDF prefixModel=GPT-OSS-20B, Feature Category=Lexical Baselines2026.05 | 0.473 | |
| TF-IDF prefixModel=R1-Distill Llama-8B, Feature Category=Lexical Baselines2026.05 | 0.47 |