The Memory Trust Gap: Capability-Dependent Failures in Persistent-Memory Agents

• URL: http://arxiv.org/abs/2609.01852v1
• Date: Tue, 01 Sep 2026 20:35:00 GMT
• Title: The Memory Trust Gap: Capability-Dependent Failures in Persistent-Memory Agents
• Abstract summary: Persistent memory supports personalized agents, but a stale stored fact can override current authoritative evidence without warning.<n>We study when this harm begins as model capability changes.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Persistent memory supports personalized agents, but a stale stored fact can override current authoritative evidence without warning. We study when this harm begins as model capability changes. We evaluate a frozen, closed-set, action-scored benchmark with 2 suites that represent 2 different meanings of "no memory" (a Benefit suite, unsolvable without the stored fact, and a Safety suite, in which an authoritative tool always holds the correct value), on a same-family model-size series (Qwen3 0.6/1.7/4/8B). The Memory Trust Gap reflects over-trust rather than confusion. In the Benefit suite, models answer with the stale value 0.92-1.00 of the time at every scale. In the Safety suite, harm below the no-memory baseline under the trap conditions ($Δ_{\mathrm{mem}}$) is capability-gated, with the larger models collapsing most once a stale note is made to look current. In a $2\times2\times2\times2$ factorial, which feature triggers over-trust depends on both the feature and model scale. Removing a label amplifies over-trust at every size, and a recency feature (stale dated newer) fools the larger models harder. Source authority is weak and scale-flat, and position changes from positive to negative across the Qwen3 model-size series. We confirm these scale interactions with direct cross-size contrast tests rather than overlapping per-model intervals. Mitigation is likewise capability-dependent: exposing metadata improves accuracy for the capable models, but only pre-resolving the conflict restores accuracy for the 2 smaller checkpoints. The same pattern appears on the capable models in an independent Llama-Instruct model-size series and on 2 external datasets (RGB, MisBench). A framing control finds no consistent advantage for the memory label: at the 3 smaller scales, models trust a stale document more than a stale memory; at 8B, the difference is not significant.

Related papers

• Calibration is the Bottleneck: An Action-Class Diagnostic of Multi-Turn Tool-Calling [65.03665015319457]
  We propose an action-class-oriented diagnostic framework that decomposes multi-turn failures into two modes: action-class miscalibration and action-execution failure.<n>We validate it on a panel of tool-calling models across multiple multi-turn benchmarks.
  arXiv  Detail & Related papers  (2026-09-01T09:08:15Z)
• When Memory Lies: An Empirical Study of Spatial Memory Staleness in VLM Agents [1.095093532533584]
  Memory-augmented VLM agents act on persistent spatial knowledge, yet that knowledge silently goes stale as the environment changes.<n>We ask what happens when an agent must reconcile a confident memory claim with a contradicting observation, and whether current models can catch the conflict before it becomes a safety-relevant mistake.
  arXiv  Detail & Related papers  (2026-08-05T08:04:26Z)
• TARL: Transaction-Aware Reliable Ledgers for Executable Memory Management in Long-Term Agents [20.123133746969398]
  We introduce TARL, a memory state update framework that maps each statement to one of five executable actions.<n> TARL identifies the affected memory, resolves its temporal scope, compares source reliability, and updates accepted, pending, and rejected ledgers.<n>We also introduce TARL-Mem, a benchmark with fine-grained action labels and next-state targets.
  arXiv  Detail & Related papers  (2026-08-04T14:02:55Z)
• Eviction as Estimation: A Fixed-Lag Smoothing View of Test-Time Memory, and When Measuring Beats Accumulating [0.0]
  A language model with a bounded working memory must repeatedly decide which stored items to keep.<n>We instantiate it as a training-free policy, RMM, a strict generalization of H2O.<n>But on independent benchmarks, run inside NVIDIA's KVPress harness against its own SnapKV, H2O, and StreamingLLM implementations, the advantage mostly disappears.
  arXiv  Detail & Related papers  (2026-07-27T17:08:27Z)
• Memory Grafting: Scaling Language Model Pre-training via Offline Conditional Memory [65.39827296429527]
  Scaling conditional memory offers a promising way to increase language-model capacity.<n>Existing methods such as Engram learn large memory tables from scratch during pre-training.<n>We propose Memory Grafting, a conditional memory scaling method that utilizes frozen hidden states from a grafting model as conditional n-gram memory.
  arXiv  Detail & Related papers  (2026-05-20T09:35:04Z)
• MemReranker: Reasoning-Aware Reranking for Agent Memory Retrieval [37.54115468116941]
  This report introduces MemReranker, a reranking model family (0.6B/4B) built on Qwen3-Reranker through multi-stage LLM knowledge distillation.<n>On the memory retrieval benchmark, MemReranker-0.6B substantially outperforms BGE-Reranker and matches open-source 4B/8B models as well as GPT-4o-mini on key metrics.<n>MemReranker-4B further achieves 0.737 MAP, with several metrics on par with Gemini-3-Flash, while maintaining inference latency at only 10-20% of large models.
  arXiv  Detail & Related papers  (2026-05-07T12:33:57Z)
• MemEvoBench: Benchmarking Memory MisEvolution in LLM Agents [78.95081012334116]
  Large Language Models (LLMs) with persistent memory enhances interaction continuity and personalization but introduces new safety risks.<n> contaminated or biased memory accumulation can trigger abnormal agent behaviors.<n>MemeEvoBench is the first benchmark evaluating memory safety in LLM agents against adversarial memory injection, noisy tool outputs, and biased feedback.
  arXiv  Detail & Related papers  (2026-04-17T07:29:52Z)
• Silent Commitment Failure in Instruction-Tuned Language Models: Evidence of Governability Divergence Across Architectures [0.0]
  We introduce governability -- the degree to which a model's errors are detectable before output commitment and correctable once detected.<n>We show benchmark accuracy does not predict governability, correction capacity varies independently of detection, and identical governance scaffolds produce opposite effects across models.<n>We propose a Detection and Correction Matrix classifying model-task combinations into four regimes: Governable, Monitor Only, Steer Blind, and Ungovernable.
  arXiv  Detail & Related papers  (2026-03-22T21:50:28Z)
• Amulet: Fast TEE-Shielded Inference for On-Device Model Protection [15.936694312917512]
  On-device machine learning (ML) introduces new security concerns about model privacy.<n> Storing valuable trained ML models on user devices exposes them to potential extraction by adversaries.<n>We propose Amulet, a fast TEE-shielded on-device inference framework for ML model protection.
  arXiv  Detail & Related papers  (2025-12-08T12:22:51Z)
• RoMa v2: Harder Better Faster Denser Feature Matching [56.71494120301684]
  Dense feature matching aims to estimate all correspondences between two images of a 3D scene.<n>Existing dense matchers fail or perform poorly for many hard real-world scenarios.<n>In this paper, we attack these weaknesses on a wide front through a series of systematic improvements.
  arXiv  Detail & Related papers  (2025-11-19T18:59:38Z)
• MICE for CATs: Model-Internal Confidence Estimation for Calibrating Agents with Tools [54.63478102768333]
  Well-calibrated model confidences can be used to weigh the risk versus reward of potential actions.<n>We propose a novel class of model-internal confidence estimators (MICE) to better assess confidence when calling tools.
  arXiv  Detail & Related papers  (2025-04-28T18:06:38Z)
• Causal Estimation of Memorisation Profiles [58.20086589761273]
  Understanding memorisation in language models has practical and societal implications. Memorisation is the causal effect of training with an instance on the model's ability to predict that instance. This paper proposes a new, principled, and efficient method to estimate memorisation based on the difference-in-differences design from econometrics.
  arXiv  Detail & Related papers  (2024-06-06T17:59:09Z)
• A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning [56.450090618578]
  Class-Incremental Learning (CIL) aims to train a model with limited memory size to meet this requirement. We show that when counting the model size into the total budget and comparing methods with aligned memory size, saving models do not consistently work. We propose a simple yet effective baseline, denoted as MEMO for Memory-efficient Expandable MOdel.
  arXiv  Detail & Related papers  (2022-05-26T08:24:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.