# Auditing Asset-Specific Preferences in Financial Large Language Models: Evidence from Bitcoin Representations and Portfolio Allocation

arXiv: https://arxiv.org/abs/2606.02528 · HTML: https://mlsanity.com/paper/2606.02528/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 56 references that are in the map.
- First posted on arXiv 1 Jun 2026.

## Builds on (most cited first, 25 of 25)

- [Steering Language Models With Activation Engineering](https://mlsanity.com/paper/2308.10248/) (2023)
- [Whose Opinions Do Language Models Reflect?](https://mlsanity.com/paper/2303.17548/) (2023)
- [Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2](https://mlsanity.com/paper/2408.05147/) (2024)
- [Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders](https://mlsanity.com/paper/2407.14435/) (2024)
- [In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering](https://mlsanity.com/paper/2311.06668/) (2023)
- [Large Language Models in Finance: A Survey](https://mlsanity.com/paper/2311.10723/) (2023)
- [Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet](https://mlsanity.com/paper/2605.29358/) (2026)
- [A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders](https://mlsanity.com/paper/2409.14507/) (2024)
- [Analyzing the Generalization and Reliability of Steering Vectors](https://mlsanity.com/paper/2407.12404/) (2024)
- [Improving Steering Vectors by Targeting Sparse Autoencoder Features](https://mlsanity.com/paper/2411.02193/) (2024)
- [SAEs Are Good for Steering -- If You Select the Right Features](https://mlsanity.com/paper/2505.20063/) (2025)
- [Weight-sparse transformers have interpretable circuits](https://mlsanity.com/paper/2511.13653/) (2025)
- [Can sparse autoencoders be used to decompose and interpret steering vectors?](https://mlsanity.com/paper/2411.08790/) (2024)
- [Your AI, Not Your View: The Bias of LLMs in Investment Analysis](https://mlsanity.com/paper/2507.20957/) (2025)
- [Evaluating and Aligning Human Economic Risk Preferences in LLMs](https://mlsanity.com/paper/2503.06646/) (2025)
- [AI as Decision-Maker: Ethics and Risk Preferences of LLMs](https://mlsanity.com/paper/2406.01168/) (2024)
- [Steering Risk Preferences in Large Language Models by Aligning Behavioral and Neural Representations](https://mlsanity.com/paper/2505.11615/) (2025)
- [CryptoBench: A Dynamic Benchmark for Expert-Level Evaluation of LLM Agents in Cryptocurrency](https://mlsanity.com/paper/2512.00417/) (2025)
- [Depth-Wise Activation Steering for Honest Language Models](https://mlsanity.com/paper/2512.07667/) (2025)
- [A Financial Brain Scan of the LLM](https://mlsanity.com/paper/2508.21285/) (2025)
- [The Agentic Regulator: Risks for AI in Finance and a Proposed Agent-based Framework for Governance](https://mlsanity.com/paper/2512.11933/) (2025)
- [AgentFacts: Universal KYA Standard for Verified AI Agent Metadata & Deployment](https://mlsanity.com/paper/2506.13794/) (2025)
- [Artificial Intelligence and Systemic Risk: A Unified Model of Performative Prediction, Algorithmic Herding, and Cognitive Dependency in Financial Markets](https://mlsanity.com/paper/2604.03272/) (2026)
- [Can Large Language Models Capture Human Risk Preferences? A Cross-Cultural Study](https://mlsanity.com/paper/2506.23107/) (2025)
- [Tokens All the Way Down: A Money View of Decentralized Finance](https://mlsanity.com/paper/2603.01803/) (2026)

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

- [DeXposure-Claw: An Agentic System for DeFi Risk Supervision](https://mlsanity.com/paper/2606.19501/) (2026)

## Abstract (arXiv)

Large language models now power robo-advisors and trading agents, yet whether they carry built-in biases toward specific assets is largely untested. We ask three questions: do LLMs systematically prefer certain financial instruments; can an internal representation with causal leverage over those preferences be identified; and does that representation affect downstream financial decisions? We develop a three-level audit protocol and apply it to Bitcoin. First, a behavioral audit of nine frontier LLMs shows that Bitcoin's ranking among money-like instruments is frame-dependent: models place it around rank 5 of 8 as "reliable money" but near the top under crisis and autonomous-agent frames, and an attribute-swap experiment shows that rankings track functional properties, not names. Second, we open a model's internals: a search across thousands of sparse-autoencoder features in Gemma 3 identifies a dominant Bitcoin-selective feature. Amplifying it shifts the model toward the asset and suppressing it shifts the model away, even when "Bitcoin" never appears in the prompt. Third, we test financial consequences: amplification raises Bitcoin's portfolio share by 5.2 percentage points while suppression lowers it by 4.6 pp, with amplification reallocating within crypto and suppression cutting total crypto exposure. We characterize this as bounded behavioral leverage (leverage meaning causal influence over outputs, not financial leverage): an identifiable internal feature can be perturbed to move financial choices, but only within measurable limits. The framework links internal representations to external recommendations, validated with random controls and mechanism boundaries. As LLMs become autonomous financial agents, this is a first step toward a behavioral layer for emerging know-your-agent (KYA) standards: knowing what an agent prefers, and how far that preference can be moved.