When Does Belief-Based Agent Memory Help? Reliability-Conditional Updating and Provenance-Capped Poisoning Defense

Abstract

We investigate when belief-based memory actually improves large language model (LLM) agents. Our vehicle is Nous, a long-term memory architecture that represents each entity-attribute pair as a categorical probability distribution updated through closed-form Bayesian inference, with information-theoretic surprise driving belief revision and entropy-based forgetting. A controlled ablation on the LoCoMo benchmark shows that Bayesian belief updating alone provides little benefit over naive last-wri

Categories

Framework mappings

Suggested from the entry's categories.

@misc{singh2026when,
  title = {{When Does Belief-Based Agent Memory Help? Reliability-Conditional Updating and Provenance-Capped Poisoning Defense}},
  author = {Pranav Singh},
  year = {2026},
  month = jun,
  eprint = {2606.22030},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.22030}
}