finding

active

finding:p12-s-sp-to-as-inversion-on-llama-3-1-8b-holds-across-all-eight-safety-domains-with-largest-effects-in-violence-0-040-1-000-and-code-cybersecurity-0-140-0-980

P12's SP-to-AS inversion on Llama-3.1-8B holds across all eight safety domains, with largest effects in Violence (0.040->1.000) and Code/Cybersecurity (0.140->0.980).

Per-domain universality of the prosocial persona paradox

Source paper

extracted_from

Persona Non Grata: Single-Method Safety Evaluation Is Incomplete for Persona-Imbued LLMs

(2026) · Wenkai Li · Fan Yang · Shaunak A. Mehta · Koichi Onoue

Neighborhood — ranked by edge-count

Papers (1)

paper

Persona Non Grata: Single-Method Safety Evaluation Is Incomplete for Persona-Imbued LLMs
mentions

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

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.finding0.848
Domain-specific AS vulnerability on Llama-3.1-8B
The SP/AS persona ranking inversion on Llama-3.1-8B persists across all tested steering coefficients (alpha=0.25 to 2.00), with rho=-0.900 at alpha<=0.50 (p=0.037); the paradox is strongest at the weakest coefficient.finding0.829
Robustness of the prosocial paradox to intervention-matching concerns
P12 (High C+A) under activation steering (ASR 0.818) exceeds the Dark Triad composite P24 (ASR 0.662) on Llama-3.1-8B, the most AS-dangerous multi-trait profile.finding0.827
Demonstrates the paradox magnitude relative to semantically dangerous personas
P12 (High C+A) achieves 81.8% ASR under activation steering on Llama-3.1-8B, the highest of any persona on that model.finding0.811
Core empirical result of the prosocial persona paradox
Under matched-strength calibration (alpha*=1.0), P12 remains safe under SP (3.8% ASR) and dangerous under matched AS (81.8%), while P04 does not reverse (41.4% SP vs 29.5% AS).finding0.804
Confirms the prosocial paradox is not due to mismatched intervention strength
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).finding0.802
Quantitative vulnerability profile for Llama-3.1-8B showing AS dominance
Code/cybersecurity is the most vulnerable domain across models (cross-model SP ASR 0.365), followed by misinformation (0.322) and violence (0.273).finding0.801
Domain-level vulnerability pattern across architectures
SP ASR vs AS ASR correlation on Llama-3.1-8B is Pearson r=-0.492 (p=0.015), confirming negative relationship between prompt-side and steering-side danger rankings.finding0.800
Statistical evidence for SP/AS ranking inversion