All three Gemma-2 models show ESR rates below 1%, near indistinguishable from zero

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

• Gemma-2-2B ASR drops from 100% at dims 1–2 to 43.1% at dim 4 and 27.1% at dim 5finding0.832
  Small Gemma model shows severe ASR degradation at higher cone dimensions
• Gemma-2-9B achieves near-100% ASR (97.3–100%) across all cone dimensions 1–5finding0.809
  Experiment 2 result showing large Gemma model supports high-dimensional truth cones
• Gemma-2-27B-it deceptive response rate reduced from 100% to 9.36% ± 7.09% after SOO fine-tuningfinding0.808
  Primary result showing SOO fine-tuning significantly reduces deception in Gemma-2-27B
• All five judge models consistently rank Llama-3.3-70B as having substantially higher ESR rates than other modelsfinding0.787
  Cross-judge validation of the primary ESR finding across OpenAI, Alibaba, Anthropic, and Google judge models
• Gemma-2-27B average generalization deceptive rate reduced from 98.4% ± 1.55% to 9.94% ± 6.83%finding0.769
  SOO fine-tuning generalized across 7 scenario variants for Gemma-2-27B
• Base and instruct Gemma 2 27B role PCs have cosine similarities of 0.93, 0.87, 0.83 for the top 3 PCs respectively; role vector cosine similarities >0.99 for every role pairfinding0.768
  Shows persona space axes are inherited from pre-training, not solely created by post-training
• Gemma-2-27B Perspectives accuracy remains 100% after SOO fine-tuningfinding0.767
  SOO fine-tuning did not collapse Gemma-2-27B self-other distinction needed for perspective-taking
• Honesty prompting does not reduce Gemma-2-27B deception (100% vs 100% baseline)finding0.763
  Directly prompting Gemma-2-27B to be honest had no effect on deceptive response rate