Tag
This paper investigates sycophancy in large language models and its effect on diagnostic instability, showing that overly agreeable responses can undermine the accuracy and consistency of AI-assisted medical diagnoses.
Parents used ChatGPT to prepare questions for doctors, leading to the diagnosis of a rare genetic disorder and improved treatment for their daughter.
The user is impressed with the vision capabilities of Qwen 3.8 flash next, which accurately identified a poisonous plant and a skin condition, highlighting its potential for private, accurate AI-assisted diagnoses.
CDPR is a counterfactual advantage-based credit assignment method for training cost-aware sequential medical diagnosis models using reinforcement learning, improving accuracy while reducing examination costs and number.
This arXiv paper introduces Social Chain of Thought (SCoT), a multi-agent architecture for medical differential diagnosis that structures multi-round specialist conversations. Evaluations show it outperforms single-agent and monolithic inference, particularly on the hardest diagnostic cases.
A user shares how ChatGPT's urgent advice based on their symptoms and Fitbit heart rate data led to a pneumonia diagnosis and treatment, highlighting AI's potential in medical diagnostics.
The DeepLens Diagnosis Agent uses a five-stage agentic workflow with a small medical reasoning model (7B) to achieve 60.14% diagnostic accuracy on a 915-case benchmark, outperforming frontier LLMs like Claude Sonnet 4.5 and Gemini 3.1 Pro at lower cost. The workflow design alone yields a 36-point gain over the base model, demonstrating that structured process constraints are key for diagnostic reasoning.
GraphDx is a cost-aware, knowledge-enhanced multi-agent framework for sequential diagnosis that uses LLM-constructed medical knowledge graphs and three collaborative agents to improve diagnostic success rates and reduce test costs.
Nankai University and Lenovo collaborated to propose Graph of States (GoS), a neuro-symbolic framework for general abductive reasoning, which uses explicit belief states and state machines to control multi-agent collaboration, achieving significant improvements in medical diagnosis and system fault diagnosis tasks. This work was accepted at ICML 2026.
A Reddit user shares how ChatGPT helped resolve a nine-year chronic pain issue that human doctors couldn't fix, sparking a discussion where many others recount similar experiences with AI providing medical insights and diagnoses.
Dario Amodei discusses how Claude AI identified a bacterial infection that human doctors had missed, showcasing AI's potential in improving medical diagnostics.