Bioalignment: Measuring and Improving LLM Disposition Toward Biological Systems for AI Safety

2026-03-09

AI Summary

This study addresses a systematic bias in large language models (LLMs) that favors non-biological solutions in technical problem-solving, potentially posing AI safety risks. We introduce the concept of “bioalignment”—the alignment of model preferences with biologically inspired or natural solutions—and present a novel evaluation benchmark comprising 50 carefully crafted prompts. Leveraging the Kelly criterion, we quantitatively assess model preferences for biological versus non-biological approaches. Through targeted intervention using QLoRA fine-tuning combined with hybrid continued pretraining and instruction tuning on a modest corpus of approximately 22 million tokens from PMC bio-inspired problem-solving texts, we significantly enhance the preference scores for biological solutions in both Llama 3.2-3B and Qwen2.5-3B (p < 0.001 and p < 0.01, respectively), without compromising general capabilities. This demonstrates that small-scale domain-specific fine-tuning can effectively recalibrate model biases toward biologically aligned reasoning.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Humans and AI: Learning Human Values and Preferences
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Large language models (LLMs) trained on internet-scale corpora can exhibit systematic biases that increase the probability of unwanted behavior. In this study, we examined potential biases towards synthetic vs. biological technological solutions across four domains (materials, energy, manufacturing, and algorithms). A sample of 5 frontier and 5 open-weight models were measured using 50 curated Bioalignment prompts with a Kelly criterion-inspired evaluation framework. According to this metric, most models were not bioaligned in that they exhibit biases in favor of synthetic (non-biological) solutions. We next examined if fine-tuning could increase the preferences of two open-weight models, Llama 3.2-3B-Instruct and Qwen2.5-3B-Instruct, for biological-based approaches. A curated corpus of ~22M tokens from 6,636 PMC articles emphasizing biological problem-solving was used first to fine-tune Llama 3B with a mixed corpus of continued training and instruction-formatted. This was then extended to Qwen 3B using instruction-formatted only. We found that QLoRA fine-tuning significantly increased the scoring of biological solutions for both models without degrading general capabilities (Holm-Bonferroni-corrected p < 0.001 and p < 0.01, respectively). This suggests that even a small amount of fine-tuning can change how models weigh the relative value of biological and bioinspired vs. synthetic approaches. Although this work focused on small open-weight LLMs, it may be extensible to much larger models and could be used to develop models that favor bio-based approaches. We release the benchmark, corpus, code, and adapter weights.

Problem

• bioalignment
• large language models
• biological systems
• synthetic bias
• AI safety

Innovation

• Bioalignment
• LLM bias mitigation
• biological preference tuning
• QLoRA fine-tuning
• AI safety evaluation