DeepLens Diagnosis Agent: Agentic Workflow Design Lets a Small Reasoning Model Compete with Frontier LLMs
DeepLens Diagnosis Agent: Agentic Workflow Design Lets a Small Reasoning Model Compete with Frontier LLMs
DeepLens 诊断智能体：智能体工作流设计让小型推理模型比肩前沿大模型
Medical diagnosis is a multi-stage process: extract facts, consult knowledge, generate a differential analysis, and select the best diagnosis with explanations. Frontier LLMs are strong generalists, but single-shot prompting often yields brittle diagnostic reasoning.
医疗诊断是一个多阶段的过程：提取事实、查阅知识、生成鉴别分析，并选择带有解释的最佳诊断结果。前沿大语言模型（LLMs）虽然是强大的通才，但单次提示（single-shot prompting）往往会导致诊断推理过程脆弱且不稳定。
We present the DeepLens Diagnosis Agent, a five-stage harnessing pipeline (combining model capabilities with disciplined process constraints) centered on a small medical reasoning model (JSL Medical Small 7B v2) and retrieval-augmented generation (RAG). The pipeline enforces structured clinical extraction, disciplined retrieval, constrained candidate generation, explicit evidence triangulation, and an auditable final decision.
我们提出了 DeepLens 诊断智能体，这是一个五阶段的利用流水线（将模型能力与严格的流程约束相结合），以小型医学推理模型（JSL Medical Small 7B v2）和检索增强生成（RAG）为核心。该流水线强制执行结构化的临床提取、规范的检索、受限的候选生成、明确的证据三角测量以及可审计的最终决策。
On the 915-case DiagnosisArena benchmark, the agent achieved 60.14% top-1 diagnostic accuracy, the highest among small and medium-sized models. The same model without the agent workflow achieved 23.99%, a +36-point gain from workflow design alone, despite 88.2% on standard medical benchmarks, showing that diagnostic reasoning under uncertainty requires more than knowledge recall.
在包含 915 个案例的 DiagnosisArena 基准测试中，该智能体实现了 60.14% 的 Top-1 诊断准确率，在中小规模模型中表现最优。同一模型在没有智能体工作流的情况下仅达到 23.99%，仅通过工作流设计就提升了 36 个百分点；尽管该模型在标准医学基准测试中表现为 88.2%，这表明在不确定性下的诊断推理不仅仅依赖于知识回忆。
The agent costs USD 0.0072 per case (24K tokens on A100) with 24-second latency, 35-45% cheaper than Claude Sonnet 4.5 (USD 0.0110) and Gemini 3.1 Pro (USD 0.0128) while outperforming them by +9.70pp and +9.17pp. Harnessing can also correct frontier model failures; workflow constraints can outweigh parameter count or API cost.
该智能体处理每个案例的成本为 0.0072 美元（在 A100 上处理 24K token），延迟为 24 秒，比 Claude Sonnet 4.5（0.0110 美元）和 Gemini 3.1 Pro（0.0128 美元）便宜 35-45%，同时性能分别领先 9.70 个百分点和 9.17 个百分点。这种利用方式还可以纠正前沿模型的错误；工作流约束的作用可以超过参数规模或 API 成本的影响。
Beyond aggregate accuracy, the pipeline produces structured intermediate artifacts that make each stage inspectable and support error localization. These properties support high-stakes settings where traceability, reproducibility, and auditable evidence matter alongside benchmark performance.
除了总体准确率外，该流水线还会产生结构化的中间产物，使每个阶段都可检查并支持错误定位。这些特性支持了高风险应用场景，在这些场景中，可追溯性、可重复性和可审计的证据与基准测试性能同样重要。