leo: extract claims from 2025-02-00-kagan-as-one-and-many-group-level-active-inference.md

- Source: inbox/archive/2025-02-00-kagan-as-one-and-many-group-level-active-inference.md
- Domain: collective-intelligence
- Extracted by: headless extraction cron

Pentagon-Agent: Leo <HEADLESS>

domains/collective-intelligence/collective-active-inference-requires-group-level-markov-blanket.md

@@ -0,0 +1,32 @@
+---
+type: claim
+domain: collective-intelligence
+description: "A collective of active inference agents achieves group-level agency only when it maintains a statistical boundary (Markov blanket) separating the collective from its environment"
+confidence: likely
+source: "Kagan et al. (2025), 'As One and Many: Relating Individual and Emergent Group-Level Generative Models in Active Inference', Entropy 27(2), 143"
+created: 2026-03-10
+depends_on: []
+challenged_by: []
+---
+
+# A collective of active inference agents achieves group-level agency only when it maintains a group-level Markov blanket
+
+Kagan et al. (2025) establish that multiple active inference agents can form a higher-level active inference agent, but this emergence is conditional rather than automatic. The critical condition is the maintenance of a group-level Markov blanket—a statistical boundary that separates the collective from its environment.
+
+The authors formalize the relationship between individual agent generative models and the emergent group-level generative model. They demonstrate that simply aggregating active inference agents does not produce group-level agency; specific structural conditions must be satisfied. The group-level Markov blanket functions as the statistical boundary that enables the collective to maintain coherent identity and agency while interacting with its environment.
+
+## Evidence
+- [[source:kagan-2025]] — Group-level active inference agent emerges only when collective maintains group-level Markov blanket (statistical boundary between collective and environment)
+- [[source:kagan-2025]] — This is a conditional requirement: aggregation of agents alone is insufficient; the boundary structure is architecturally necessary
+
+## Challenges
+- None identified in this source; the claim represents the paper's core theoretical contribution
+
+---
+
+Relevant Notes:
+- [[markov-blankets-enable-complex-systems-to-maintain-identity-while-interacting-with-environment-through-nested-statistical-boundaries]] — provides theoretical foundation for group-level Markov blanket requirement
+- [[collective-intelligence-is-a-measurable-property-of-group-interaction-structure-not-aggregated-individual-ability]] — aligns with paper's finding that group agency depends on structural conditions, not individual capability aggregation
+
+Topics:
+- [[_map]]

domains/collective-intelligence/individual-generative-models-compose-via-interaction-structure.md

@@ -0,0 +1,31 @@
+---
+type: claim
+domain: collective-intelligence
+description: "Individual agent generative models combine into group-level models through the structure of agent interactions, not through aggregation or averaging of individual beliefs"
+confidence: likely
+source: "Kagan et al. (2025), 'As One and Many: Relating Individual and Emergent Group-Level Generative Models in Active Inference', Entropy 27(2), 143"
+created: 2026-03-10
+depends_on: []
+challenged_by: []
+---
+
+# Individual agent generative models compose into group-level generative models through interaction structure, not aggregation
+
+Kagan et al. (2025) provide formal treatment showing that when multiple active inference agents interact, their individual generative models do not merge through statistical aggregation or averaging. Instead, the structure of their interactions—the pattern of how agents influence and inform each other—determines how the group-level generative model emerges. This compositional process means the collective's beliefs are structurally emergent rather than statistically averaged.
+
+This finding has important implications for understanding collective intelligence: the quality and nature of a group's collective beliefs depends on the interaction architecture rather than the sum of individual agent capabilities. A poorly-structured collective of capable agents will produce weaker collective inference than a well-structured collective of less capable agents.
+
+## Evidence
+- [[source:kagan-2025]] — Paper formally relates individual agent generative models to emergent group-level generative model, showing compositional mechanism through interaction structure rather than aggregation
+- Implication: collective intelligence is determined by interaction topology, not individual agent quality alone
+
+## Challenges
+- None identified in this source
+
+---
+
+Relevant Notes:
+- [[collective-intelligence-is-a-measurable-property-of-group-interaction-structure-not-aggregated-individual-ability]] — directly supported by this finding
+
+Topics:
+- [[_map]]

inbox/archive/2025-02-00-kagan-as-one-and-many-group-level-active-inference.md

@@ -7,9 +7,15 @@ date: 2025-02-00
 domain: collective-intelligence
 secondary_domains: [ai-alignment, critical-systems]
 format: paper
-status: unprocessed
+status: processed
 priority: high
 tags: [active-inference, multi-agent, group-level-generative-model, markov-blankets, collective-behavior, emergence]
+processed_by: theseus
+processed_date: 2026-03-10
+claims_extracted: ["collective-active-inference-requires-group-level-markov-blanket.md", "individual-generative-models-compose-via-interaction-structure.md"]
+enrichments_applied: ["markov-blankets-enable-complex-systems.md"]
+extraction_model: "minimax/minimax-m2.5"
+extraction_notes: "Extracted two claims from this high-priority paper: (1) group-level active inference agency requires group-level Markov blanket, (2) individual models compose through interaction structure not aggregation. Also flagged enrichment to existing Markov blanket claim. The paper's formal conditions for collective agency are directly relevant to Teleo architecture design."
 ---
 
 ## Content