Image de couverture de livre

Are LLM Belief Updates Consistent with Bayes' Theorem?

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Imran, Sohaib, Kendiukhov, Ihor, Broerman, Matthew, Thomas, Aditya, Campanella, Riccardo, Lamb, Rob, Atkinson, Peter M.
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Out-of-Context Abduction: LLMs Make Inferences About Procedural Data Leveraging Declarative Facts in Earlier Training Data
par: Imran, Sohaib, et autres
Publié: (2025)

Consistency Training while Mitigating Obfuscation via Rate Matching
par: Imran, Sohaib, et autres
Publié: (2026)

Systematic Evaluation of Single-Cell Foundation Model Interpretability Reveals Attention Captures Co-Expression Rather Than Unique Regulatory Signal
par: Kendiukhov, Ihor
Publié: (2026)

Multi-Dimensional Spectral Geometry of Biological Knowledge in Single-Cell Transformer Representations
par: Kendiukhov, Ihor
Publié: (2026)

A Review of Developmental Interpretability in Large Language Models
par: Kendiukhov, Ihor
Publié: (2025)

PABU: Progress-Aware Belief Update for Efficient LLM Agents
par: Jiang, Haitao, et autres
Publié: (2026)

VAL-Bench: Belief Consistency as a measure for Value Alignment in Language Models
par: Gupta, Aman, et autres
Publié: (2025)

Steamroller Problems: An Evaluation of LLM Reasoning Capability with Automated Theorem Prover Strategies
par: McGinness, Lachlan, et autres
Publié: (2024)

The Belief State Transformer
par: Hu, Edward S., et autres
Publié: (2024)

DeepTheorem: Advancing LLM Reasoning for Theorem Proving Through Natural Language and Reinforcement Learning
par: Zhang, Ziyin, et autres
Publié: (2025)

Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs?
par: Hase, Peter, et autres
Publié: (2024)

Opir: Efficient Multi-Task Safety Classification for Toxicity, Jailbreaks, Hate Speech, and Harmful Content
par: Stepanov, Ihor, et autres
Publié: (2026)

FairBelief -- Assessing Harmful Beliefs in Language Models
par: Setzu, Mattia, et autres
Publié: (2024)

Ergodicity Library: A Python Toolkit for Stochastic-Process Simulation, Time-Average Diagnostics, and Agent-Based Experiments
par: Kendiukhov, Ihor
Publié: (2026)

Preference-Aware Memory Update for Long-Term LLM Agents
par: Sun, Haoran, et autres
Publié: (2025)

Mining the Mind: What 100M Beliefs Reveal About Frontier LLM Knowledge
par: Ghosh, Shrestha, et autres
Publié: (2025)

Faithful and Robust LLM-Driven Theorem Proving for NLI Explanations
par: Quan, Xin, et autres
Publié: (2025)

Accumulating Context Changes the Beliefs of Language Models
par: Geng, Jiayi, et autres
Publié: (2025)

Belief-Sim: Towards Belief-Driven Simulation of Demographic Misinformation Susceptibility
par: Borah, Angana, et autres
Publié: (2026)

Harnessing Consistency for Robust Test-Time LLM Ensemble
par: Zeng, Zhichen, et autres
Publié: (2025)

AXCEL: Automated eXplainable Consistency Evaluation using LLMs
par: Sreekar, P Aditya, et autres
Publié: (2024)

Trustworthy LLM-Mediated Communication: Evaluating Information Fidelity in LLM as a Communicator (LAAC) Framework in Multiple Application Domains
par: Rafi, Mohammed Musthafa, et autres
Publié: (2025)

Resource-Efficient Fine-Tuning of LLaMA-3.2-3B for Medical Chain-of-Thought Reasoning
par: Mansha, Imran
Publié: (2025)

ConsistencyChecker: Tree-based Evaluation of LLM Generalization Capabilities
par: Hong, Zhaochen, et autres
Publié: (2025)

TheoremExplainAgent: Towards Video-based Multimodal Explanations for LLM Theorem Understanding
par: Ku, Max, et autres
Publié: (2025)

BaRDa: A Belief and Reasoning Dataset that Separates Factual Accuracy and Reasoning Ability
par: Clark, Peter, et autres
Publié: (2023)

Simulating LLM-to-LLM Tutoring for Multilingual Math Feedback
par: Tonga, Junior Cedric, et autres
Publié: (2025)

CSCE: Boosting LLM Reasoning by Simultaneous Enhancing of Causal Significance and Consistency
par: Wang, Kangsheng, et autres
Publié: (2024)

TriAlign: Towards Universal Truth Consistency in Personalized LLM Alignment
par: Nguyen, Thi-Nhung, et autres
Publié: (2026)

GraphMind: Theorem Selection and Conclusion Generation Framework with Dynamic GNN for LLM Reasoning
par: Li, Yutong, et autres
Publié: (2025)

Curved Inference: Concern-Sensitive Geometry in Large Language Model Residual Streams
par: Manson, Rob
Publié: (2025)

Vulnerability of LLMs' Stated Beliefs? LLMs Belief Resistance Check Through Strategic Persuasive Conversation Interventions
par: Huang, Fan, et autres
Publié: (2026)

Narrative Theory-Driven LLM Methods for Automatic Story Generation and Understanding: A Survey
par: Liu, David Y., et autres
Publié: (2026)

ABBEL: LLM Agents Acting through Belief Bottlenecks Expressed in Language
par: Lidayan, Aly, et autres
Publié: (2025)

GLiNER multi-task: Generalist Lightweight Model for Various Information Extraction Tasks
par: Stepanov, Ihor, et autres
Publié: (2024)

Mind the Gap: How Elicitation Protocols Shape the Stated-Revealed Preference Gap in Language Models
par: Mahajan, Pranav, et autres
Publié: (2026)

Line of Duty: Evaluating LLM Self-Knowledge via Consistency in Feasibility Boundaries
par: Kale, Sahil, et autres
Publié: (2025)

Evaluating Consistencies in LLM responses through a Semantic Clustering of Question Answering
par: Lee, Yanggyu, et autres
Publié: (2024)

Reasoning Aware Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling
par: Wan, Guangya, et autres
Publié: (2024)

DecMetrics: Structured Claim Decomposition Scoring for Factually Consistent LLM Outputs
par: Huang, Minghui
Publié: (2025)