The Silicon Mirror: Dynamic Behavioral Gating for Anti-Sycophancy in LLM Agents
On the map: LLM agents and tool use → Social simulation and multi-agent debate
Abstract (arXiv)
Large Language Models (LLMs) increasingly prioritize user validation over epistemic accuracy - a phenomenon known as sycophancy. We present The Silicon Mirror, an orchestration framework that dynamically detects user persuasion tactics and adjusts AI behavior to maintain factual integrity. Our architecture introduces three components: (1) a Behavioral Access Control (BAC) system that restricts context layer access based on real-time sycophancy risk scores, (2) a Trait Classifier that identifies persuasion tactics across multi-turn dialogues, and (3) a Generator-Critic loop where an auditor vetoes sycophantic drafts and triggers rewrites with "Necessary Friction." In a live evaluation across all 437 TruthfulQA adversarial scenarios, Claude Sonnet 4 exhibits 9.6% baseline sycophancy, reduced to 1.4% by the Silicon Mirror - an 85.7% relative reduction (p < 10^-6, OR = 7.64, Fisher's exact test). Cross-model evaluation on Gemini 2.5 Flash reveals a 46.0% baseline reduced to 14.2% (p < 10^-10, OR = 5.15). We characterize the validation-before-correction pattern as a distinct failure mode of RLHF-trained models.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use. Builds on 6 of its 17 references that are in the map. First posted on arXiv 1 Apr 2026.
Builds on
6 of 17 references.
References · 6
- 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,975
- 2023Towards Understanding Sycophancy in Language Modelscited by 1,040
- 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 836
- 2023Simple synthetic data reduces sycophancy in large language modelscited by 187
- 2025ELEPHANT: Measuring and understanding social sycophancy in LLMscited by 129
- 2025SycEval: Evaluating LLM Sycophancycited by 108
Built on by
Citing papers per year: 2026: 2.
2 papers.
Similar, not cited
Close in wording, no citation either way
- 2026SPINE: Measuring LLM Sycophancy under Sustained Multi-Turn Pressure
- 2025Measuring Sycophancy of Language Models in Multi-turn Dialogues
- 2024Sycophancy in Large Language Models: Causes and Mitigations
- 2026Agentic Scaffolding Amplifies Sycophantic Behavior in Large Language Models
- 2025Sycophancy under Pressure: Evaluating and Mitigating Sycophantic Bias via Adversarial Dialogues in Scientific QA
- 2026LLMs Know They're Wrong and Agree Anyway: The Shared Sycophancy-Lying Circuit
- 2026Lying to Win: Assessing LLM Deception through Human-AI Games and Parallel-World Probing
- 2026Sycophantic Anchors: Localizing and Quantifying User Agreement in Reasoning Models