# Nous: Learning and Certifying Memory Decisions Before Source Calibration

arXiv: https://arxiv.org/abs/2610.00094 · HTML: https://mlsanity.com/paper/2610.00094/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 29 references that are in the map.
- First posted on arXiv 8 Sep 2026.

## Builds on (most cited first, 16 of 16)

- [MultiWOZ -- A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling](https://mlsanity.com/paper/1810.00278/) (2018)
- [A Reductions Approach to Fair Classification](https://mlsanity.com/paper/1803.02453/) (2018)
- [MultiWOZ 2.4: A Multi-Domain Task-Oriented Dialogue Dataset with Essential Annotation Corrections to Improve State Tracking Evaluation](https://mlsanity.com/paper/2104.00773/) (2021)
- [Classification with Asymmetric Label Noise: Consistency and Maximal Denoising](https://mlsanity.com/paper/1303.1208/) (2013)
- [Fairness Evaluation in Presence of Biased Noisy Labels](https://mlsanity.com/paper/2003.13808/) (2020)
- [MemTX: Transactional Belief Commit for Stateful Agent Memory](https://mlsanity.com/paper/2607.23929/) (2026)
- [Classification with unknown class-conditional label noise on non-compact feature spaces](https://mlsanity.com/paper/1902.05627/) (2019)
- [Learning from Label Proportion with Online Pseudo-Label Decision by Regret Minimization](https://mlsanity.com/paper/2302.08947/) (2023)
- [Decontamination of Mutual Contamination Models](https://mlsanity.com/paper/1710.01167/) (2017)
- [MemTxn: A Transaction Boundary for Source-Supported Updates and Complete-State Recovery in Agent Memory](https://mlsanity.com/paper/2607.27834/) (2026)
- [Can Agent Memory Systems Track Evolving State?](https://mlsanity.com/paper/2608.19652/) (2026)
- [When Does Belief-Based Agent Memory Help? Reliability-Conditional Updating and Provenance-Capped Poisoning Defense](https://mlsanity.com/paper/2606.22030/) (2026)
- [Weak Supervision Performance Evaluation via Partial Identification](https://mlsanity.com/paper/2312.04601/) (2023)
- [Fundamental limits for learning hidden Markov model parameters](https://mlsanity.com/paper/2106.12936/) (2021)
- [Noise tolerance of learning to rank under class-conditional label noise](https://mlsanity.com/paper/2208.02126/) (2022)
- [Predictive Performance Comparison of Decision Policies Under Confounding](https://mlsanity.com/paper/2404.00848/) (2024)

## Abstract (arXiv)

Agent memory systems update state decisions from reports whose reliability may be unknown. Existing analyses of source estimation do not determine when a policy can be learned or its improvement certified without identifying the reporting channel. We study these three tasks using the same observed records. For a specified hidden Markov family with continuous source uncertainty, decision learning and powered certification have quadratic sample complexity, whereas fixed-precision source estimation has quartic complexity. We characterize a sharp identified interval for policy gain under an unknown shared-background channel and derive finite-sample certificates under bounded history dependence and conditional copying. Independently trained witness regions support general history spaces, and disagreement-conditioned auditing improves power for sparse revisions. For dependent histories, prediction-count-preserving batches cancel the unknown reporting background and admit conditional certificates. A MultiWOZ 2.4 evaluation uses text-processing policies on 1,000 human-written test dialogues with simulated audits. Balanced batches retain 2.26 percentage points of the full candidate's 7.34 percentage-point mean gain and obtain more positive certificates under weak audits. These results establish task-specific information requirements and provide an auditable policy-revision framework for Nous.