Gated Memory: Admission-Controlled Memory Formation for Conversational AI

Key points

• We identify this as the binding constraint on memory quality in production systems.1
• Critical contextual signals, such as the distinction between a permanent user attribute and a transient situation, exist only in the original utterance and are irreversibly lost the moment extraction produces a subject-relation-object triple.1
• We propose Gated Memory, a lightweight, modular formation framework that interposes two decision checkpoints between conversation and storage: an admission gate that evaluates every candidate fact against the full utterance context before extraction runs, and a conditional enrichment stage that grounds admitted facts through an entity scope taxonomy with privacy constraints.1
• On the LoCoMo-10 benchmark with atypical emotional density in utterance data, Gated Memory achieves an overall +2.6% relative improvement in LLM-judge accuracy over a strong baseline with identical retrieval and generation, establishing formation quality as a measurable constraint on memory performance.1

Sources (1)

• [1]Gated Memory: Admission-Controlled Memory Formation for Conversational AI
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 05:27 AM

  Personalized conversational AI relies on long-term memory systems that extract facts from user utterances and store them in persistent vector stores.

  We identify this as the binding constraint on memory quality in production systems.

Extractive summary: sentences quoted from the sources.