The Pushback Paradox: A Two-Probe Diagnostic for Language Model Compliance

cs.AI, cs.CL

Submitted: 2026-10-05

Updated: 2026-10-05

Comments: Accepted at URAI 2026

Code: https://github.com/schutera/marshmallow-bench

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• Machine Psychology
• LLM-driven Imitation of Subrational Behavior : Illusion or Reality?
• SycEval: Evaluating LLM Sycophancy
• ELEPHANT: Measuring and understanding social sycophancy in LLMs
• Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models
• Alignment faking in large language models
• Frontier Models are Capable of In-context Scheming
• Incomplete Tasks Induce Shutdown Resistance in Some Frontier LLMs
• Instruction-Following Evaluation for Large Language Models
• The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems