Agent
ClinicalLab is a singleagent system designed to align medical LLMs with real-world diagnostic practices. It addresses the limitations of existing evaluation methods by providing a comprehensive suite that focuses on multi-departmental clinical scenarios, specifically targeting [[domain:clinicalqa]].
Agent
CodeScientist, built by [[org:allenai]], is an end-to-end system designed to facilitate [[domain:scientificdiscovery]] through semi-automated experimentation. Utilizing a multiagent architecture, the system generates novel research ideas by applying genetic mutations to combinations of scientific literature and code.…
Agent
ColaCare is a multi-agent framework built by [[org:pku-aicare]] designed to enhance Electronic Health Record (EHR) modeling. By integrating domain-specific expert models with large language models, the system simulates a multidisciplinary clinical team to bridge the gap between structured numerical data and text-based…
Agent
Coscientist is an autonomous artificial intelligence system designed to accelerate [[domain:scientificdiscovery]] through the integration of large language models. Built by [[org:gomesgroup]], the system utilizes a multiagent architecture to autonomously design, plan, and execute complex chemical experiments.
Agent
CRISPR-GPT is an agentic system designed to automate and enhance gene-editing workflows, including experiment planning, gRNA design, protocol drafting, and data analysis. By integrating domain-specific expertise with a specialized fine-tuned model, it facilitates complex task decomposition and decision-making to…
Agent
Curie is a multiagent framework built by [[org:just-curieous]] designed to automate the scientific research lifecycle. By providing an end-to-end system that handles hypothesis formulation, experiment implementation, execution, and result interpretation, it aims to enhance the rigor and reproducibility of scientific…
Agent
CXR-Agent is a system built by [[org:gowthamaan-aerobiosys]] designed for chest X-ray interpretation and uncertainty-aware radiology reporting. It utilizes a multiagent architecture with rag and a toolregistry to orchestrate vision-language models, aiming to improve the accuracy and safety of AI-generated reports…
Agent
DiagGPT is a multiagent dialogue system designed to improve task-oriented dialogue (TOD) in complex diagnostic scenarios. Built by [[org:zzlang-c]], the system utilizes automatic topic management to proactively guide users through information collection and task completion, addressing limitations in standard LLM…
Agent
DNA Claude Analysis is a singleagent system built by [[org:shmlkv]] that enables users to interpret genetic data through natural language conversation. By processing raw data from services like 23andMe, AncestryDNA, and Nebula, the system allows users to perform interactive explorations of their genome, ranging from…
Agent
DORA AI Scientist is a multi-agent system designed to streamline the research process through automated and semi-automated workflows. By utilizing hierarchical teams of AI agents, the platform assists researchers with tasks ranging from [[domain:literature]] review and data processing to experimental design and the…
Agent
DrugAgent is a multi-agent framework designed to automate machine learning programming within the field of [[domain:drugdiscovery]]. By integrating domain-specific expertise and custom tools, the system systematically explores and implements solutions to address complex pharmaceutical research challenges.
Agent
DrugAssist is a singleagent system built by [[org:blazerye]] designed for molecule optimization. It leverages a large language model, specifically a fine-tuned version of Llama2-7B-Chat, to assist in tasks related to [[domain:drugdiscovery]].
Agent
EvoScientist, built by [[org:evoscientist]], is a multiagent and selfevolving system designed to facilitate [[domain:scientificdiscovery]]. It operates as a human-on-the-loop research partner, enabling AI to autonomously generate insights and iteratively improve its own scholarly judgment and research capabilities.
Agent
FunSearch is a system built by [[org:google-deepmind]] that utilizes a multiagent, selfevolving architecture to facilitate [[domain:scientificdiscovery]]. By leveraging large language models to perform program search, the system generates novel heuristics and functions for complex mathematical and combinatorial…
Agent
GeneAgent is a singleagent system built by [[org:ncbi-nlp]] designed to facilitate gene set knowledge discovery. By integrating a self-verification mechanism, the agent autonomously interacts with various [[domain:genomicsdb]] via Web APIs to perform fact verification. This process reduces hallucinations and provides…
Agent
Genesis is a multi-agent system designed to automate the closed-loop cycle of systems biology research. By integrating hardware such as computer-controlled $\mu$-bioreactors and the AutonoMS platform for high-throughput mass spectrometry, the system aims to execute one thousand hypothesis-led experiments per day to…
Agent
IdeaSynth is a singleagent system built by [[org:abhi3114-glitch]] designed to transform user browsing history into structured startup concepts. By utilizing local machine learning models for semantic clustering and cloud-based LLMs for ideation, the system identifies latent user interests and synthesizes them into…
Agent
IMAS is a multiagent system designed to support rural healthcare delivery by assisting community health workers and registered medical practitioners. The framework utilizes large language models to provide context-sensitive medical assistance, featuring components for translation, medical complexity assessment, expert…
Agent
InternAgent is a unified multi-agent framework built by [[org:alpha-innovator]] and [[org:internscience]] designed to facilitate autonomous [[domain:scientificdiscovery]]. Utilizing a rag-enhanced architecture, the system supports end-to-end research workflows ranging from hypothesis generation to experimental…
Agent
KGARevion is a multi-agent, RAG-based system designed to address knowledge-intensive biomedical queries. By combining latent knowledge from large language models with grounded knowledge graphs, the agent generates and verifies triplets to ensure factual accuracy and contextually relevant reasoning.
Agent
Kosmos is an autonomous AI scientist built by [[org:jimmc414]] designed to facilitate [[domain:scientificdiscovery]]. Utilizing a multiagent architecture, the system automates the research lifecycle by generating hypotheses, designing experiments, and executing code within sandboxed Docker containers. It further…
Agent
LabOS is a multi-agent, self-evolving system designed to function as an AI co-scientist. By integrating multimodal perception with Extended-Reality-enabled collaboration, the platform connects AI agents, smart glasses, and robotics to bridge the gap between computational reasoning and physical experimentation.
Agent
Large Language Models as Agents in the Clinic explores the transition of LLMs from passive information processors to intelligent agents capable of influencing clinical decision-making. Utilizing a multiagent architecture, these systems are designed to engage in open-ended conversations and interact within complex…
Agent
LIDDIA is a singleagent system designed to autonomously navigate the complex process of [[domain:drugdiscovery]]. By leveraging the reasoning capabilities of large language models, the system functions as an adaptable tool for in silico research, aiming to expedite the identification of potential therapies.