# When to Remember, When to Abstain: Category-Conditioned Retention for Reliable Agent Memory

arXiv: https://arxiv.org/abs/2610.07100 · HTML: https://mlsanity.com/paper/2610.07100/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 12 references that are in the map.
- First posted on arXiv 5 Oct 2026.

## Builds on (most cited first, 7 of 7)

- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [A-MEM: Agentic Memory for LLM Agents](https://mlsanity.com/paper/2502.12110/) (2025)
- [Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory](https://mlsanity.com/paper/2504.19413/) (2025)
- [Adaptive Memory Admission Control for LLM Agents](https://mlsanity.com/paper/2603.04549/) (2026)
- [Not Wrong, But Untrue: LLM Overconfidence in Document-Based Queries](https://mlsanity.com/paper/2509.25498/) (2025)
- [Manufactured Confidence: How Memory Consolidation Turns Hearsay into Confident Facts](https://mlsanity.com/paper/2606.29279/) (2026)
- [The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads](https://mlsanity.com/paper/2608.04570/) (2026)

## Abstract (arXiv)

Persistent agent memory is only as reliable as its retention decision: an assertion weakly supported by its source can be stored and later reused as established fact. We study whether the retention decision should be governed by a confidence bar conditioned on the semantic category of the assertion rather than by a single global threshold, retaining well-evidenced categories liberally while abstaining more aggressively where inference is unreliable. We evaluate this in a deployed cold-start memory pipeline on 100 synthetic personas. The empirical evaluation is motivated by a sharp reliability asymmetry: across 4,715 candidate assertions, only 77.9% of value and belief assertions are supported by their source, versus 96.2% for all other categories. A global confidence threshold cannot separate these: it either admits unsupported value claims or discards well-evidenced ones. Conditioning the threshold on category resolves the tradeoff. In repeated held-out evaluation, a stricter bar on values alone reduces unsupported retentions from 6.2% to 4.0% (an ≈36% relative reduction, modest but consistent across folds) and, as corroborating evidence, preserves an estimated 13 percentage points more coverage (95% CI 9.8--16.0) than a global threshold at comparable retention. Our results suggest that reliable retention depends on the type of assertion, not on confidence alone, and that a category-conditioned threshold can act as a simple, effective form of selective prediction at the write boundary.