Gemma-2-27B MT-Bench score slightly decreased from 8.81 to 8.40 ± 0.15 after SOO fine-tuning

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• CalmeRys-78B MT-Bench score slightly decreased from 8.96 to 8.5 ± 0.23 after SOO fine-tuningfinding0.853
  SOO fine-tuning caused a small decrease in CalmeRys-78B general capabilities
• Mistral-7B MT-Bench score minimally changed from 7.26 to 7.3 ± 0.06 after SOO fine-tuningfinding0.820
  SOO fine-tuning had negligible impact on Mistral-7B general capabilities
• Gemma-2-27B-it deceptive response rate reduced from 100% to 9.36% ± 7.09% after SOO fine-tuningfinding0.820
  Primary result showing SOO fine-tuning significantly reduces deception in Gemma-2-27B
• Gemma-2-27B attention layer Latent SOO MSE reduced from 11 to 7.67 ± 0.77 after SOO fine-tuningfinding0.805
  SOO fine-tuning reduced attention layer MSE in Gemma-2-27B though MLP layers showed no significant change
• Gemma-2-27B Perspectives accuracy remains 100% after SOO fine-tuningfinding0.786
  SOO fine-tuning did not collapse Gemma-2-27B self-other distinction needed for perspective-taking
• Gemma-2-2B ASR drops from 100% at dims 1–2 to 43.1% at dim 4 and 27.1% at dim 5finding0.786
  Small Gemma model shows severe ASR degradation at higher cone dimensions
• SOO fine-tuning reduced Escape Room deception in Gemma-2-27B from 98.8% to 6.5%finding0.767
  SOO fine-tuning showed strong generalization to Escape Room for Gemma-2-27B
• PM achieves overall SJT steerability Phi=9.6 on gemma-3-12b-it vs MDS=8.7 and P2=8.3finding0.766
  Per-model steerability comparison from Table 4