lighteternal/biodecision-tev1-4b
BioDecision-4B by lighteternal is a 4.5 billion parameter System-1 decision model based on Qwen3.5-4B, specifically designed for biomedicine, pharma, and clinical trials. It processes a source, question, and predefined answers to return calibrated probabilities for each option in a single forward pass. This model excels at tasks like clinical-trial outcome forecasting, patient-trial matching, and judging AI-written answers against a source, demonstrating significant performance gains over general-purpose models in these specialized biomedical domains.
Loading preview...
BioDecision-4B: A Specialized Biomedical Decision Model
BioDecision-4B, developed by lighteternal, is a 4.5 billion parameter model built upon the Qwen3.5-4B architecture, fine-tuned as a System-1 decision engine for biomedical applications. Unlike general-purpose LLMs, it is optimized to read a source text, a question, and a set of possible answers, then output a calibrated probability for each answer in a single pass. This design allows for efficient and precise decision-making across various biomedical tasks without requiring retraining for new decision types.
Key Capabilities and Performance
- Biomedical Decision-Making: Achieves 70.2% accuracy across 46,199 held-out decisions from 28 biomedical benchmarks, outperforming Qwen3.5-4B and Together's Tev1-4B by over 7 points.
- Clinical Trial Forecasting: Demonstrates strong performance in clinical-trial outcome forecasting (CT Open Endpoint), with macro-F1 scores of 72.6 and 62.0, comparable to or exceeding models like GPT-5 and Claude Opus 4.5.
- Patient-Trial Matching: Significantly improves NDCG@10 to 0.833 for patient-trial matching (TREC 2022).
- AI Answer Grounding: Excels at judging AI-written answers against a source, achieving 87.9% on HaluBench PubMedQA, on par with GPT-4.1.
- Efficiency: Processes decisions rapidly, with a latency of 44 ms per decision and a throughput of 93 decisions/s on an A100 GPU, leading to low operational costs.
Differentiators and Use Cases
This model's primary differentiation lies in its specialized focus and high accuracy for specific biomedical decision tasks, where it often surpasses larger, more generalist models. It is particularly well-suited for:
- Automated assessment of clinical trial outcomes.
- Efficiently matching patients to clinical trials.
- Validating the factual accuracy of AI-generated content in biomedical contexts.
- Relation extraction in biomedical texts.
It is important to note that BioDecision-4B is designed for research use and is not validated for patient care. It focuses on decision-making from provided sources and options, without built-in retrieval or explanation generation capabilities.