Belief Propagation

Follow this topic Ask about this topic

Latest papers

14 papers4 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

Where Is the Truth When AI Agents Disagree? Dependence-Aware Belief Propagation for Unsupervised Truth Discovery (Extended Version)
Nov 3, 2026Laure Berti-EquilleLLM Agent Reliability Multi-Agent Consensus
IRD, ESPACE-DEV, Montpellier, France
Joint End-to-End Training of Weighted BP Decoder Ensembles
Oct 24, 2026Felix Krieg, Paul Bezner, Stephan ten BrinkError-Correcting Codes Belief Propagation
Institute of Telecommunications, University of Stuttgart, Germany
On the Adversarial Vulnerability of Neural Receivers with Unrolled LDPC Decoders
Oct 21, 2026Aakash Saini, Jafar Aco Mohammadi, Wolfgang GerstackerAdversarial Attacks Neural Network Robustness
Nokia, Munich, Germany · Institute for Digital Communications, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
Slow Beats Fast at the Kesten-Stigum Threshold: Minimax, Fisher-Information and Belief-Propagation Characterizations of the Information-Computation Gap in Sparse Stochastic Block Models
Oct 6, 2026Soroor GhandaliBelief Propagation Low-Degree Polynomial Method
A theory of platonic representations in language models
Oct 5, 2026Darshil Doshi, Wenjie Zhou, Corinna Elena Wegner +3Belief Propagation Multilingual Language Models
Johns Hopkins Univeristy · Johns Hopkins University & EPFL
Hierarchy-GBP: Accelerating Factor Graph Inference via Abstraction and Recovery
Oct 4, 2026Yuzhou Cheng, Tom Yates, Ignacio Alzugaray +3Gaussian Belief Propagation Belief Propagation
Imperial College London, London, United Kingdom.
Spike-based Belief Propagation in Nonlinear Dynamical Systems
Aug 20, 2026Sepideh Adamiat, Hongye Wang, Wouter M. Kouw +1Belief Propagation Adaptive Control
Electrical Engineering Department, Eindhoven University of Technology, Eindhoven, the Netherlands · Biomedical Signals and Systems, Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS), University of Twente, Enschede, The Netherlands
Collective Ranking of Environmental Signals through Gaussian Belief Propagation in a Patrolling Robot Swarm
Aug 18, 2026Zachary R. Madin, Connor York, Jonathan Lawry +1Gaussian Belief Propagation Multi-Robot Systems
University of Bristol, Bristol, UK
Information Geometry of Message Passing
Aug 16, 2026Mykola Lukashchuk, Kyrylo Yemets, Alex Ledbetter +1Belief Propagation Probabilistic Graphical Models
Eindhoven University of Technology · Lviv Polytechnic National University · Lazy Dynamics, Utrecht, the Netherlands
Reduction of Probabilistic Chemical Reaction Networks
Jun 26, 2026Mauricio Montes, Gregoire Sergeant-PerthuisBelief Propagation Probabilistic Graphical Models
COSAM, Auburn University, Auburn, AL, USA · CQSB, Sorbonne Universite, Paris, France
Equivariant Neural Belief Propagation
Jun 4, 2026Zehua Cheng, Wei Dai, Jiahao SunEquivariant Representation Learning Factor Graph Optimization
Department of Computer Science, University of Oxford, Oxford, United Kingdom · FLock.io, London, United Kingdom
Exact Regular-Constrained Variable-Order Markov Generation via Sparse Context-State Belief Propagation
May 8, 2026François PachetConstrained Generative Modeling Belief Propagation
LIP6, Sorbonne Université · Ynosound
Learning Markov Processes as Sum-of-Square Forms for Analytical Belief Propagation
Apr 8, 2026Peter Amorese, Morteza LahijanianBelief Propagation Markov Models
University of Colorado Boulder
Transformers are Bayesian Networks
Mar 17, 2026Gregory CoppolaBelief Propagation Transformer

Nov 3, 2026J/K move · Enter open · S save

Where Is the Truth When AI Agents Disagree? Dependence-Aware Belief Propagation for Unsupervised Truth Discovery (Extended Version)

Laure Berti-Equille

IRD, ESPACE-DEV, Montpellier, France

LLM Agent Reliability Multi-Agent Consensus Belief Propagation

Open paper AskPDF arXiv

LLM agents often give conflicting answers to the same question, and many deployed agents are variants of one base model, so their errors are correlated. We study truth discovery among such agents: inferring the true answer and each agent's reliability without labels. Classical truth-discovery and crowdsourcing estimators reward an agreeing bloc and can fall below majority vote. We propose BP-TD, which adds per-agent decorrelation weights to mean-field variational inference on a one-coin Dawid-Skene model, so that agents sharing a base model count together as about one vote. Its first iteration is exactly a weighted vote; later iterations estimate agent reliability. With 13 local agents on MMLU, BP-TD beats majority vote without any labels (0.716 vs. 0.624 on 160 questions; 0.707 vs. 0.674 on 10,000) and ranks agents by their true accuracy, which classical estimators invert. The gain depends on the pool: over 2,852 sub-pools of four benchmarks, decorrelation helps when the largest correlated bloc is weaker than the other agents and hurts when it is the strongest, and we found no label-free statistic that tells the two cases apart. Labels change the picture quickly: with 20 labelled questions, selecting the single best agent is more accurate than every aggregator we tested on MMLU, ARC and TruthfulQA.