{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "title": "Mem0: incrementally selecting and updating conversational memories", "summary": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses.", "boundary": "These are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.", "area": "External conversational memory", "topic": "Artificial intelligence", "year": 2025, "lane": "HUMAN", "result": "ALGORITHM", "package_id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "humans": [ "Prateek Chhikara", "Dev Khant", "Saket Aryan", "Taranjeet Singh", "Deshraj Yadav" ], "authors_complete": true, "human_role_note": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261017-arxiv-2504-19413v1", "title": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1", "date": "2025-04-28" } ], "points": "50000", "published_grade": { "id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "entry_id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "band": "B", "modifier": "+", "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "what_changed": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses.", "scope": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses. These are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-23T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Prateek Chhikara", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Dev Khant", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Saket Aryan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Taranjeet Singh", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Deshraj Yadav", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "These are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.", "domain": "External conversational memory" }, "sources": [ { "citation": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1" } ], "statement": "# Mem0: incrementally selecting and updating conversational memories\n\n## A useful correction must reach the memory\n\nSuppose a public project first plans a Friday release and later postpones it to Monday. Saving both sentences without their relationship can make an assistant answer with the old date. The 2025 Mem0 paper describes a pipeline that extracts possible memories, finds related stored items and decides whether to add, update, delete or ignore a candidate. The graph variant also represents relationships between entities.\n\n## Four actions, several possible mistakes\n\nAdding an item preserves something new. Updating changes an existing memory. Deleting removes one, while ignoring leaves the store unchanged. A model chooses among these actions from the available context. It can select the wrong old item, lose an exception or mistake a hypothetical plan for a confirmed change. The operations are a design for managing memory, not an independent guarantee that every stored statement is true.\n\n## Read the denominator in a headline\n\nIf a classroom baseline scores fifty points and a method scores sixty-five, the difference is fifteen percentage points on that scale. The relative increase is fifteen divided by fifty, or thirty percent. Neither number says how the method compares with a different baseline scoring seventy. These invented numbers explain why the original paper’s relative gains must be read against the named comparison, model and question set.\n\n## Make a correction trail\n\nAsk your agent to process three public project notes: the initial plan, the postponement and a later conditional proposal. Require a table showing the old item, proposed action, source and remaining uncertainty. Keep the original notes available even when making a simplified memory view. The paper reports lower context use and latency alongside mixed answer scores; those measurements are not all current product capabilities or total lifecycle costs.\n\n## Separate memory edits from source edits\n\nFor the three-note exercise, save a view showing the current plan and a small table explaining why an older plan was superseded. Keep the original notes intact. This practice is a source-audit aid, not a requirement that every commercial memory system use the same history policy as WRS.\n\n## Scientific scope\n\nThese are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.\n\n## Sources for this explanation\n\n- [Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory](https://arxiv.org/abs/2504.19413v1): Original 2025 method and benchmark tables: extraction/update operations, graph variant, full-context comparison, manual memory proxy and stage-specific token/latency measurements.\n", "summary": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses.", "title": "Mem0: incrementally selecting and updating conversational memories" }, "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "source_ref": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:3379678aaa691a7eb52899197c06c41dac0762a10773b18de487b05f75f2ad5f" }, "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "entry_id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "revision": 1, "title": "Mem0: incrementally selecting and updating conversational memories", "context": "External conversational memory", "track": "Reusable scientific method, experimental result or historical framework", "band": "B", "modifier": "+", "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-23", "evidence_cutoff": "2026-10-23", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses.", "scope": "A long-running assistant may need a few relevant facts rather than an entire conversation history. The 2025 Mem0 paper describes extracting candidate memories, finding related stored items and deciding whether to add, update, delete or ignore each candidate. A graph variant also stores relationships. The researchers report conversational-memory results and query-time costs. The contribution is the specified incremental memory pipeline and its source-bound evaluation, rather than a certification of the current commercial product or ownership of the models it uses. These are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.", "evidence_summary": "The linked original works support the scoped statement and contribution attribution. WRS has not independently replayed experiments or reconstructed a full mathematical proof. These are author/vendor-reported experiments. Full-context inference has a higher overall judge score in the reported table, while Mem0 uses less retrieved context and has lower measured latency. A headline relative gain compares against a particular manual memory proxy, not every current OpenAI system. Extraction and updating are not included in every cost figure. Adversarial questions are excluded from the main comparison, and a quoted competing result has a category-label discrepancy that prevents an unqualified cross-paper ranking.", "references": [ { "title": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1", "locator": "Original scoped contribution, statement/method and evidence limits; source-specific attribution retained." } ], "roles": [ { "who": "The named authors in the research record", "credit": "SCIENCE", "did": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred." } ], "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "50000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": { "id": "chhikara-et-al-mem0-incremental-conversational-memory-2025", "target": { "record_id": "urn:uuid:bbc2ee92-8d03-4909-a454-135b3b92fcfb", "version": 1, "content_hash": "sha256:805ef6c7e55e75ab851f3c157223c96c4be5234d7ea22a0c240750385c7228ae" }, "sections": [ { "title": "A useful correction must reach the memory", "text": "Suppose a public project first plans a Friday release and later postpones it to Monday. Saving both sentences without their relationship can make an assistant answer with the old date. The 2025 Mem0 paper describes a pipeline that extracts possible memories, finds related stored items and decides whether to add, update, delete or ignore a candidate. The graph variant also represents relationships between entities." }, { "title": "Four actions, several possible mistakes", "text": "Adding an item preserves something new. Updating changes an existing memory. Deleting removes one, while ignoring leaves the store unchanged. A model chooses among these actions from the available context. It can select the wrong old item, lose an exception or mistake a hypothetical plan for a confirmed change. The operations are a design for managing memory, not an independent guarantee that every stored statement is true." }, { "title": "Read the denominator in a headline", "text": "If a classroom baseline scores fifty points and a method scores sixty-five, the difference is fifteen percentage points on that scale. The relative increase is fifteen divided by fifty, or thirty percent. Neither number says how the method compares with a different baseline scoring seventy. These invented numbers explain why the original paper’s relative gains must be read against the named comparison, model and question set." }, { "title": "Make a correction trail", "text": "Ask your agent to process three public project notes: the initial plan, the postponement and a later conditional proposal. Require a table showing the old item, proposed action, source and remaining uncertainty. Keep the original notes available even when making a simplified memory view. The paper reports lower context use and latency alongside mixed answer scores; those measurements are not all current product capabilities or total lifecycle costs." }, { "title": "Separate memory edits from source edits", "text": "For the three-note exercise, save a view showing the current plan and a small table explaining why an older plan was superseded. Keep the original notes intact. This practice is a source-audit aid, not a requirement that every commercial memory system use the same history policy as WRS." } ], "references": [ { "title": "Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory", "url": "https://arxiv.org/abs/2504.19413v1", "locator": "Original 2025 method and benchmark tables: extraction/update operations, graph variant, full-context comparison, manual memory proxy and stage-specific token/latency measurements." } ] }, "reading_status": "EXPLANATORY_GUIDE" }