GEST
Authoring for Living Worlds: Tool-Constrained LLM Agents for Executable Multi-Actor Scenarios
Nicolae Cudlenco, Mihai Masala, Marius Leordeanu · 2026-07-29
- Management
- T2 Dynamic Memory
- Memory changes during deployment through a stable management policy.
- Representation
- Cross-represented explicit
How experience becomes memory
Each path is one memory component, traced from what is stored to how it is found again.
Persistent simulated world state via the event graph (entities/state across events)
- Represented asSpatial, temporal & cross-modal structured memoryCross-represented explicit
- Stored inGraph / relational store
- Retrieved byStructure traversal, Planner / tool-assisted retrieval
- Source evidenceSource not retained
Executable event-graph nodes; World state. The same entities and state are carried across events in the simulated world.Full-paper method / memory representation and retrieval · PDF p. 2 · Figure 1 / method overview
How memory changes (T2 operations)
Update · Rule-based update
During operation, GEST applies fixed-rule insertion or replacement to Persistent simulated world state via the event graph (entities/state across events); the memory state changes while the governing update policy remains fixed.
Revision · Rule-based state correction
During operation, GEST applies fixed-rule correction of stored state to Persistent simulated world state via the event graph (entities/state across events); the memory state changes while the governing update policy remains fixed.
Reorganization · Relinking & reindexing
During operation, GEST applies relinking or reindexing of stored items to Persistent simulated world state via the event graph (entities/state across events); the memory state changes while the governing update policy remains fixed.
Related systems
- AniMET2 · Video & streaming
- AViLAT2 · Video & streaming
- CANVAST2 · Video & streaming
- DigimonGPTT2 · Video & streaming
- EGAgentT2 · Video & streaming
- EventMemAgentT2 · Video & streaming