LLaMA-3-8B-Instruct achieves highest ACCatom = 0.65 among evaluated models

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

• Llama-3.1-8B-Instructconcept0.826
  Primary qualitative demonstration model and one of 14 LLMs benchmarked
• Llama-3.3-70B-Instructconcept0.816
  Primary model of interest showing substantial ESR; largest model tested in the study
• LLaMA-3.1-8B: Sbmax = -1.896 ± 0.211, AUSN = -2.119 ± 0.198, peak layer ℓ* = 10 (median)finding0.814
  Seed-pooled geometry-only statistics (per-dev z units).
• 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).finding0.810
  Quantitative vulnerability profile for Llama-3.1-8B showing AS dominance
• Llama-3.3-70B shows multi-attempt rate of 7.4% vs. ≤1.2% for all other models testedfinding0.808
  Supporting finding showing ESR is driven by both higher multi-attempt rates and comparable improvement rates
• Llama-3.2-3B-Instructconcept0.800
  3B Llama model tested; used for injection stride visualization
• Llama-2-7b achieves only 17% harmless and 12% helpful HH-intent scores at baseline.finding0.796
  Numerical result from Table 3 showing smallest Llama model performance.
• Meta-Llama-3.1-8B-Instructconcept0.792
  Backbone model used in E3 geometry analysis.