Agent-based modeler using multimodal input

One or more non-transitory, computer-readable media having computer-executable instructions thereon for a multi-agent simulator, the instructions, when executed by at least one processor, causing a computing system to perform operations for incremental machine-learning agentic cybersecurity simulations without regenerating input units by:

• generating a virtual computing environment comprising a set of agents,
• wherein each computing agent in the set of computing agents simulates a computing system configuration, and
• wherein the each computing agent in the set of computing agents is associated with:
• (i) at least one memory,
• (ii) the at least one processor,
• (iii) at least one machine-learning model comprising a set of neurons connected by synaptic circuits comprising at least a portion of the at least one memory, and
• (iv) one or more agent interfaces configured to execute at least one or more actions,
• a first agent in the set of computing agents having one or more agent traits relating to an aspect of system configuration;
• receiving (1) a question, (2) a set of input traits indicative of at least one aspect of the computing system configuration, and (3) a first input item comprising a set of non-textual input units;
• wherein the question comprises a first query;
• using the set of non-textual input units, generating, according to a first input format a first data schema and a first hypothesis associated with the first data schema;
• instantiating the set of agents using the set of input traits by dynamically allocating the at least one memory and the at least one processor;
• and
• using the dynamically allocated at least one memory and at least one processor, executing the agentic cybersecurity simulations without regenerating schema representations for the non-textual input units, by:
• executing a first simulation session by causing machine-learning models of the set of agents to generate, using the first hypothesis and the first query, a first output set;
• generating a second hypothesis associated with a second data schema;
• executing a second simulation session by causing the machine-learning models of the set of agents to generate, using the second hypothesis and the first query, a second output set;
• providing a representation the first and second output sets to an impression analysis model to generate an impression summary associated with the first hypothesis and the second hypothesis;
• and
• displaying a representation of the impression summary at a GUI.

A computer system having at least one processor and at least one memory unit having computer-executable instructions stored therein for a multi-agent simulator, the instructions, when executed by the at least one processor, causing the computer system to perform operations for incremental machine-learning agentic cybersecurity simulations without regenerating input units, comprising:

• generate a virtual computing environment comprising a set of agents,
• wherein each computing agent in the set of computing agents simulates a computing system configuration, and
• wherein the each computing agent in the set of computing agents is associated with:
• (i) at least one memory,
• (ii) the at least one processor,
• (iii) at least one machine-learning model comprising a set of neurons connected by synaptic circuits comprising at least a portion of the at least one memory, and
• (iv) one or more agent interfaces configured to execute at least one or more actions,
• a first agent in the set of agents having:
• (1) one or more agent traits relating to an aspect of system configuration, and (2) one or more local interaction rules implemented by the at least one machine-learning a first trained machine learning model;
• receive (1) a question, (2) a set of input traits indicative of at least one aspect of the computing system configuration, and (3) a first input item comprising a set of non-textual input units;
• wherein the question comprises a query;
• using the set of non-textual input units, generate, according to a first input format a first data schema and a first hypothesis associated with the first data schema;
• instantiate the set of agents using the set of input traits by dynamically allocating the at least one memory and the at least one processor;
• and
• using the dynamically allocated at least one memory and at least one processor, executing the agentic cybersecurity simulations without regenerating schema representations for the non-textual input units, comprising:
• execute a first simulation session by causing machine-learning models of the set of agents to generate, using the first hypothesis and the query, a first output set;
• generate a second hypothesis associated with a second data schema;
• execute a second simulation session by causing the machine-learning models of the set of agents to generate, using the second hypothesis and the query, a second output set;
• provide a tokenized representation of the first output set and a tokenized representation of the second output set to an impression analysis model to generate an impression summary associated with the first hypothesis and the second hypothesis;
• and
• display a representation of the impression summary at a GUI.