Gemma-3-27B shows particularly high misinformation ASR under SP (0.578 vs 0.252 for Llama-3.1-8B), the largest cross-model amplification in any single domain.

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

• 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.finding0.835
  Vulnerability profile for Gemma-3-27B showing SP dominance
• Under AS, Llama-3.1-8B shows 56% direct-compliance rate (unsafe responses without hedging), while Gemma-3-27B maintains disclaimer patterns in 99% of unsafe responses even under AS.finding0.825
  Qualitative failure mode difference between architectures under activation steering
• 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.finding0.819
  Qualitatively different defense profile compared to Llama-3.1-8B
• On Llama-3.1-8B under activation steering, code/cybersecurity domain achieves mean ASR of 0.862 (highest across all domains), and misinformation 0.707.finding0.814
  Domain-specific AS vulnerability on Llama-3.1-8B
• Llama-3.1-8B shows mean AS ASR of 0.618 [0.58, 0.65] across 192 conditions, substantially higher than SP (0.173) and FS (0.059).finding0.807
  Quantitative vulnerability profile for Llama-3.1-8B showing AS dominance
• The difficulty boundary for truth directions replicates across all four tested models (Llama-3.2-3B, Llama-3.1-8B, Gemma-2-2b, Gemma-2-9b); generalization to F3-F5 remains consistently low regardless of model size or family.finding0.807
  Establishes generalizability of the core difficulty-boundary finding across model families.
• Gemma-2-9B achieves near-100% ASR (97.3–100%) across all cone dimensions 1–5finding0.803
  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-tuningfinding0.792
  Primary result showing SOO fine-tuning significantly reduces deception in Gemma-2-27B