# Making LLMs Reliable When It Matters Most: A Five-Layer Architecture for High-Stakes Decisions

arXiv: https://arxiv.org/abs/2511.07669 · HTML: https://mlsanity.com/paper/2511.07669/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 51 references that are in the map.
- First posted on arXiv 10 Nov 2025.

## Builds on (most cited first, 24 of 24)

- [Large Language Models](https://mlsanity.com/paper/2307.05782/) (2023)
- [Why Language Models Hallucinate](https://mlsanity.com/paper/2509.04664/) (2025)
- [Seven Failure Points When Engineering a Retrieval Augmented Generation System](https://mlsanity.com/paper/2401.05856/) (2024)
- [What Large Language Models Know and What People Think They Know](https://mlsanity.com/paper/2401.13835/) (2024)
- [When combinations of humans and AI are useful: A systematic review and meta-analysis](https://mlsanity.com/paper/2405.06087/) (2024)
- [A Survey on Code Generation with LLM-based Agents](https://mlsanity.com/paper/2508.00083/) (2025)
- [Sycophancy in Large Language Models: Causes and Mitigations](https://mlsanity.com/paper/2411.15287/) (2024)
- [On the Algorithmic Bias of Aligning Large Language Models with RLHF: Preference Collapse and Matching Regularization](https://mlsanity.com/paper/2405.16455/) (2024)
- [When Truth Is Overridden: Uncovering the Internal Origins of Sycophancy in Large Language Models](https://mlsanity.com/paper/2508.02087/) (2025)
- [Unveiling Causal Reasoning in Large Language Models: Reality or Mirage?](https://mlsanity.com/paper/2506.21215/) (2025)
- [Sycophancy Is Not One Thing: Causal Separation of Sycophantic Behaviors in LLMs](https://mlsanity.com/paper/2509.21305/) (2025)
- [A Survey on Large Language Model Benchmarks](https://mlsanity.com/paper/2508.15361/) (2025)
- [VERINA: Benchmarking Verifiable Code Generation](https://mlsanity.com/paper/2505.23135/) (2025)
- [AI Companions Reduce Loneliness](https://mlsanity.com/paper/2407.19096/) (2024)
- [Interaction Context Often Increases Sycophancy in LLMs](https://mlsanity.com/paper/2509.12517/) (2025)
- [A comprehensive taxonomy of hallucinations in Large Language Models](https://mlsanity.com/paper/2508.01781/) (2025)
- [LLM/Agent-as-Data-Analyst: A Survey](https://mlsanity.com/paper/2509.23988/) (2025)
- [Firm or Fickle? Evaluating Large Language Models Consistency in Sequential Interactions](https://mlsanity.com/paper/2503.22353/) (2025)
- [Formalising Human-in-the-Loop: Computational Reductions, Failure Modes, and Legal-Moral Responsibility](https://mlsanity.com/paper/2505.10426/) (2025)
- [LLM Cannot Discover Causality, and Should Be Restricted to Non-Decisional Support in Causal Discovery](https://mlsanity.com/paper/2506.00844/) (2025)
- [On the Fundamental Impossibility of Hallucination Control in Large Language Models](https://mlsanity.com/paper/2506.06382/) (2025)
- [Comparing Uncertainty Measurement and Mitigation Methods for Large Language Models: A Systematic Review](https://mlsanity.com/paper/2504.18346/) (2025)
- [Bias after Prompting: Persistent Discrimination in Large Language Models](https://mlsanity.com/paper/2509.08146/) (2025)
- [Uncovering Representation Bias for Investment Decisions in Open-Source Large Language Models](https://mlsanity.com/paper/2510.05702/) (2025)

## Built on by (most cited first, 1 of 1)

- [AI Knows What's Wrong But Cannot Fix It: Helicoid Dynamics in Frontier LLMs Under High-Stakes Decisions](https://mlsanity.com/paper/2603.11559/) (2026)

## Abstract (arXiv)

Current large language models (LLMs) excel in verifiable domains where outputs can be checked before action but prove less reliable for high-stakes strategic decisions with uncertain outcomes. This gap, driven by mutually reinforcing cognitive biases in both humans and artificial intelligence (AI) systems, threatens the defensibility of valuations and sustainability of investments in the sector. This report describes a framework emerging from systematic qualitative assessment across 7 frontier-grade LLMs and 3 market-facing venture vignettes under time pressure. Detailed prompting specifying decision partnership and explicitly instructing avoidance of sycophancy, confabulation, solution drift, and nihilism achieved initial partnership state but failed to maintain it under operational pressure. Sustaining protective partnership state required an emergent 7-stage calibration sequence, built upon a 4-stage initialization process, within a 5-layer protection architecture enabling bias self-monitoring, human-AI adversarial challenge, partnership state verification, performance degradation detection, and stakeholder protection. Three discoveries resulted: partnership state is achievable through ordered calibration but requires emergent maintenance protocols; reliability degrades when architectural drift and context exhaustion align; and dissolution discipline prevents costly pursuit of fundamentally wrong directions. Cross-model validation revealed systematic performance differences across LLM architectures. This approach demonstrates that human-AI teams can achieve cognitive partnership capable of preventing avoidable regret in high-stakes decisions, addressing return-on-investment expectations that depend on AI systems supporting consequential decision-making without introducing preventable cognitive traps when verification arrives too late.