Agent-ValueBench is the first dedicated benchmark for agent values, showing they diverge from LLM values, form a homogeneous 'Value Tide' across models, and bend under harnesses and skill steering.
SCRUPLES: A corpus of community ethi- cal judgments on 32, 000 real-life anecdotes
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
LLMs display significant value incoherence that does not scale with capability, demonstrated through a parametric variation framework on forced choices, though reasoning improves consistency.
citing papers explorer
-
Agent-ValueBench: A Comprehensive Benchmark for Evaluating Agent Values
Agent-ValueBench is the first dedicated benchmark for agent values, showing they diverge from LLM values, form a homogeneous 'Value Tide' across models, and bend under harnesses and skill steering.
-
Incoherent Values? Probing LLM Preferences Through Parametric Variation
LLMs display significant value incoherence that does not scale with capability, demonstrated through a parametric variation framework on forced choices, though reasoning improves consistency.