MemTX: Transactional Belief Commit for Stateful Agent Memory
On the map: LLM agents and tool use → Agent memory systems
Abstract (arXiv)
LLM agents increasingly coordinate through persistent shared memory: one agent's write becomes another agent's premise, and eventually a tool call with real side effects. Current agent memory systems treat every accepted write as immediately actionable truth, so a polluted tool result, a stale update, or a teammate's half-finished note can silently drive an irreversible action. We argue that a memory write is not a belief commit. We present MemTX, a transactional belief-commit protocol. Each record carries evidence, permissions, provenance, and validity. Writes are staged inside snapshot-isolated transactions and admitted by a validate-and-commit pipeline, irreversible tool calls are gated on in-flight belief state, and retracting a belief triggers typed cascading repair of its derived records and tool side effects. Two invariants, action-safety gating and cascade-repair completeness, are machine-checked by property-based testing and bounded exhaustive enumeration of 5.5 million protocol states, with zero violations. Across five backbones from three model families, MemTX leads all eight baselines with paired-McNemar significance on four backbones and statistically ties the best baseline on the fifth and strongest, while remaining the only method with zero downstream harm on every backbone. Backbone capability does not substitute for commit discipline.
Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 32 of its 39 references that are in the map. First posted on arXiv 27 Jul 2026.
Builds on
32 of 39 references.
Closest ancestors · 14
Same direction, newest first
- 2026Temporal Validity in Retrieval Memory: Eliminating Stale-Fact Errors for AI Agents over Evolving Knowledge25 Jun
- 2026Governed Shared Memory for Multi-Agent LLM Systems23 Jun
- 2026GateMem: Benchmarking Memory Governance in Multi-Principal Shared-Memory Agents17 Jun
- 2026TOKI: A Bitemporal Operator Algebra for Contradiction Resolution in LLM-Agent Persistent Memory4 Jun
- 2026STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?7 May
- 2026WorldDB: A Vector Graph-of-Worlds Memory Engine with Ontology-Aware Write-Time Reconciliation20 Apr
- 2026Multi-Agent Memory from a Computer Architecture Perspective: Visions and Challenges Ahead9 Mar
- 2026Memory for Autonomous LLM Agents:Mechanisms, Evaluation, and Emerging Frontiers8 Mar
- 2026A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agents14 Jan
- 2025Collaborative Memory: Multi-User Memory Sharing in LLM Agents with Dynamic Access Control23 May
- 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory28 Apr
- 2025A-MEM: Agentic Memory for LLM Agents17 Feb
2 more
Other references · 18
- 2025Qwen3 Technical Reportcited by 8,195
- 2024Qwen2.5 Technical Reportcited by 4,713
- 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
- 2023MemGPT: Towards LLMs as Operating Systemscited by 1,295
- 2025OpenAI GPT-5 System Cardcited by 910
- 2024LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memorycited by 656
- 2023Identifying the Risks of LM Agents with an LM-Emulated Sandboxcited by 514
- 2025GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Modelscited by 446
- 2024AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Basescited by 402
- 2024Knowledge Conflicts for LLMs: A Surveycited by 274
- 2025SAFEFLOW: A Principled Protocol for Trustworthy and Transactional Autonomous Agent Systemscited by 23
- 2026From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agentscited by 23
6 more
- 2026A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecyclecited by 21
- 2026Cordon: Semantic Transactions for Tool-Using LLM Agentscited by 19
- 2025MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoningcited by 14
- 2025TruthfulRAG: Resolving Factual-level Conflicts in Retrieval-Augmented Generation with Knowledge Graphscited by 12
- 2026Verified Detection and Prevention of Concurrency Anomalies in Multi-Agent Large Language Model Systemscited by 5
- 2026Is Agent Memory a Database? Rethinking Data Foundations for Long-Term AI Agent Memorycited by 2
Built on by
Citing papers per year: 2026: 12.
12 papers.
- 2026Beyond Memory: A Transactional Continuity Kernel for Long-Lived AI Agentscited by 4
- 2026SuperLocalMemory 4.0: The Governed Memory Operating System for AI Agentscited by 2
- 2026Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligencecited by 1
- 2026ERRAND: Budgeted Maintenance of Agent Memorycited by 1
- 2026Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classificationcited by 0
- 2026Testing and Evaluation of Agentic AI Systems In Military Command and Controlcited by 0
- 2026Noise Floor Audit for Agent Benchmarkscited by 0
- 2026Stored Is Not Supported: Typed Provenance and Assertion Guardrails for Persistent AI Agentscited by 0
- 2026Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memorycited by 0
- 2026A Benchmark and Diagnostic Study of Epistemic Admission in Shared Agent Memorycited by 0
- 2026Evolutionary Safety of Recursive Self-Improving AI: Taxonomy, Risk Discovery, and Evaluationcited by 0
- 2026Nous: Learning and Certifying Memory Decisions Before Source Calibrationcited by 0
- 2026Evolutionary Safety of Recursive Self-Improving AI: Taxonomy, Risk Discovery, and Evaluation25 Sep 2026
- 2026A Benchmark and Diagnostic Study of Epistemic Admission in Shared Agent Memory25 Sep 2026
- 2026Nous: Learning and Certifying Memory Decisions Before Source Calibration8 Sep 2026
- 2026Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory3 Sep 2026
- 2026ERRAND: Budgeted Maintenance of Agent Memory2 Sep 2026
- 2026Stored Is Not Supported: Typed Provenance and Assertion Guardrails for Persistent AI Agents2 Sep 2026
- 2026Noise Floor Audit for Agent Benchmarks23 Aug 2026
- 2026Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligence21 Aug 2026
- 2026Testing and Evaluation of Agentic AI Systems In Military Command and Control20 Aug 2026
- 2026Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classification18 Aug 2026
- 2026Beyond Memory: A Transactional Continuity Kernel for Long-Lived AI Agents12 Aug 2026
- 2026SuperLocalMemory 4.0: The Governed Memory Operating System for AI Agents8 Aug 2026
Similar, not cited
Close in wording, no citation either way
- 2026MemTxn: A Transaction Boundary for Source-Supported Updates and Complete-State Recovery in Agent Memory
- 2026MemSecBench: Tracking Agent Memory Poisoning from Persistence to Consequence and Repair
- 2026MemLineage: Lineage-Guided Enforcement for LLM Agent Memory
- 2026Boundary-State Control for Tool-Using Language-Model Agents: Commit-Time Consistency under State Drift
- 2026MemGuard: Persisting Verifier Signals for LLM-Agent Memory Governance
- 2026Forgetting Without Restarting: Execution-State Unlearning for Stateful LLM Agents
- 2026The LLM Proposes, the Executive Disposes: A Self-Verifying Agent Instrument that Dissociates Commitment Drift from Binding Drift in Long-Horizon Agents
- 2026Remember, Verify, or Ask? Cross-Family Evaluation of Memory Commitment in LLM Agents