gemma-3-1b-it yields only one valid MDS injection score (phi_1,A,up = 4.8) and is excluded from main analyses

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

• Why do MDS injections fail on gemma-3-1b-it but succeed across all other tested LLMs?question0.814
  Unexplained exception identified as a limitation and open question
• PM achieves overall SJT steerability Phi=9.6 on gemma-3-12b-it vs MDS=8.7 and P2=8.3finding0.781
  Per-model steerability comparison from Table 4
• Gemma-3-27B and Qwen3.5-27B show near-uniform AS vulnerability regardless of persona identity (AS ASR range 0.095-0.117 on Gemma-3-27B, 0.015-0.052 on Qwen3.5-27B), indicating safety mechanisms robust to geometric persona perturbation.finding0.774
  Qualitatively different defense profile compared to Llama-3.1-8B
• Gemma-3-27B shows mean SP ASR of 0.316 [0.30, 0.33] and AS ASR of only 0.108 [0.09, 0.13], showing SP >> AS vulnerability pattern.finding0.773
  Vulnerability profile for Gemma-3-27B showing SP dominance
• Gemma-2-27B-it deceptive response rate reduced from 100% to 9.36% ± 7.09% after SOO fine-tuningfinding0.767
  Primary result showing SOO fine-tuning significantly reduces deception in Gemma-2-27B
• Gemma-2-2B ASR drops from 100% at dims 1–2 to 43.1% at dim 4 and 27.1% at dim 5finding0.766
  Small Gemma model shows severe ASR degradation at higher cone dimensions
• MDS injections outperform P2 in open-ended generation in 11 of 14 LLMs with Phi gains of 3.61% to 16.44%finding0.764
  Primary quantitative result overturning prior reports that prompting outperforms representation engineering
• Gemma 3 4B wellbeing probe: peak Cohen's d=1.8finding0.758
  Weaker cross-family probe; explains weaker introspection in Gemma