🤖 AI Summary
Current large language models struggle to capture the underlying causal mechanisms when predicting clinical trial outcomes. To address this limitation, this work proposes a novel training paradigm that integrates supervised fine-tuning, reinforcement learning guided by validation-based rewards, and synthetic counterfactual reasoning trajectories. By leveraging counterfactual pairs to construct training data, the approach steers models—particularly those under 10B parameters, such as Qwen3.5-9B—toward learning interpretable biomedical causal reasoning processes through iterative counterfactual imagination. The method substantially outperforms both unadapted language models and conventional correlation-based baselines, while simultaneously generating transparent and human-interpretable reasoning pathways that reflect plausible causal structures in clinical contexts.
📝 Abstract
Predicting the outcomes of prospective clinical trials remains a major challenge for large language models. Prior work has shown that both traditional correlational predictors, such as random forests and logistic regression, and strong commercial LLMs achieve limited performance on this task. In this paper, we propose DeepImagine, a framework for teaching LLMs biomedical reasoning through successive counterfactual imagining. The central idea is to approximate hidden causal mechanisms of clinical trials by training models to infer how observed trial results would change under controlled perturbations of experimental conditions, such as dosage, outcome measures, study arms, geography, and other trial attributes. To support this objective, we construct both natural and approximate counterfactual pairs from real clinical trials with reported outcomes. For settings where strict counterfactual supervision is available, such as paired outcome measures or dose-ranging study arms within the same trial, we train models with supervised fine-tuning. For broader settings where only approximate counterfactual pairs can be retrieved, we optimize models with reinforcement learning using verifiable rewards based on downstream benchmark correctness. We further augment training with synthetic reasoning traces that provide causally plausible explanations for local counterfactual transitions. Using this pipeline, we train language models under 10B parameters, including Qwen3.5-9B, and evaluate them on clinical trial outcome prediction. We aim to show that DeepImagine consistently improves over untuned language models and traditional correlational baselines. Finally, we aim to show that the learned reasoning trajectories provide interpretable signals about how models represent trial-level mechanisms, suggesting a practical path toward more mechanistic and scientifically useful biomedical language models.