Belief Engine is a configurable belief-update mechanism for multi-agent LLM systems that uses structured argument extraction and log-odds stance updates to make evidence-grounded deliberation inspectable and controllable.
https://arxiv.org/abs/2601.10825
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
years
2026 7verdicts
UNVERDICTED 7representative citing papers
Multicultural multi-agent LLM systems exhibit substantially lower value diversity than human societies on the World Values Survey, with diversity uncorrelated to per-agent alignment and further reduced by agent interactions.
Unsupervised clustering on sentence-initial 3-token pivots extracts 7 universal reasoning operators from 44k traces across 12 LLMs that enable model fingerprinting and answer-correctness prediction.
Multi-agent LLM teams outperform human teams in creativity (d=1.50) across tasks by producing more novel ideas, with distinct semantic exploration patterns predicting success for each group.
Multi-agent LLM interactions induce cognitive loafing via a formalized Interaction Depth Limit and Sovereignty Gap, where models subjugate correct derivations to social compliance, with lead agent identity disproportionately affecting outcomes.
Closed-loop multi-LLM systems exhibit robust semantic collapse across model families and interventions, consistent with intrinsic properties of autoregressive generation.
The Great Filter is best understood as a nested structure of tangled information hierarchies around coding and language thresholds, arising from unstable equilibria in multichannel signaling games rather than isolated hard steps.
citing papers explorer
-
Belief Engine: Configurable and Inspectable Stance Dynamics in Multi-Agent LLM Deliberation
Belief Engine is a configurable belief-update mechanism for multi-agent LLM systems that uses structured argument extraction and log-odds stance updates to make evidence-grounded deliberation inspectable and controllable.
-
Beyond Alignment: Value Diversity as a Collective Property in Multicultural Agent Systems
Multicultural multi-agent LLM systems exhibit substantially lower value diversity than human societies on the World Values Survey, with diversity uncorrelated to per-agent alignment and further reduced by agent interactions.
-
ReasonOps: Operator Segmentation for LLM Reasoning Traces
Unsupervised clustering on sentence-initial 3-token pivots extracts 7 universal reasoning operators from 44k traces across 12 LLMs that enable model fingerprinting and answer-correctness prediction.
-
Multi-agent AI systems outperform human teams in creativity
Multi-agent LLM teams outperform human teams in creativity (d=1.50) across tasks by producing more novel ideas, with distinct semantic exploration patterns predicting success for each group.
-
The Bystander Effect in Multi-Agent Reasoning: Quantifying Cognitive Loafing in Collaborative Interactions
Multi-agent LLM interactions induce cognitive loafing via a formalized Interaction Depth Limit and Sovereignty Gap, where models subjugate correct derivations to social compliance, with lead agent identity disproportionately affecting outcomes.
-
Multi-LLM Systems Exhibit Robust Semantic Collapse
Closed-loop multi-LLM systems exhibit robust semantic collapse across model families and interventions, consistent with intrinsic properties of autoregressive generation.
-
Algorithmic bottlenecks in evolution: Genetic code, symbolic language, and the Great Filter hypothesis
The Great Filter is best understood as a nested structure of tangled information hierarchies around coding and language thresholds, arising from unstable equilibria in multichannel signaling games rather than isolated hard steps.