{"slug": "profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through", "title": "Profile-Graph Memory for LLM Agents: Implicit Cross-Entity Traversal through Narrative Profiles", "summary": "Researchers introduce MemHop, a multi-hop memory benchmark of 1,000 questions at hop depths 1-5 across 10 social-network scenarios, and ProGraph, a two-layer memory architecture combining profile expansion and compression residuals, achieving 80.1% on MemHop and 78.4% on LoCoMo, outperforming Mem0, A-Mem, HippoRAG, and RAG.", "body_md": "arXiv:2607.19359v1 Announce Type: new\nAbstract: Long-term memory is essential for LLM agents that interact across sessions, yet current memory benchmarks primarily evaluate single-hop recall, leaving multi-hop association largely unmeasured. We make three contributions. First, we introduce MemHop, a multi-hop memory benchmark of 1,000 questions at hop depths 1-5 across 10 social-network scenarios, with per-hop evidence annotations. Second, we present Profile-Graph Memory (ProGraph), a two-layer memory architecture combining (i) profile expansion -- substring-matched traversal of entity names that naturally appear in LLM-written profile narratives, a minimal alternative to explicit knowledge-graph construction -- and (ii) compression residuals -- exact dates, quantities, and named items co-extracted with each profile update at zero extra API cost. Third, a full-grid ablation shows cross-benchmark mechanism specialization: profile expansion drives multi-hop reasoning (-22.6pp on MemHop when removed) while compression residuals drive precision recall (-8.6pp on LoCoMo when not co-extracted), with cross-effects under 3pp within a single architecture. ProGraph averages 80.1% on MemHop (matching the FullContext reference) and 78.4% on LoCoMo (exceeding FullContext by 11.3pp), outperforming Mem0, A-Mem, HippoRAG, and RAG on both. We release MemHop, ProGraph, and baseline implementations.", "url": "https://wpnews.pro/news/profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through", "canonical_source": "https://arxiv.org/abs/2607.19359", "published_at": "2026-07-23 04:00:00+00:00", "updated_at": "2026-07-23 04:08:39.036855+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-research"], "entities": ["MemHop", "ProGraph", "LoCoMo", "Mem0", "A-Mem", "HippoRAG", "RAG"], "alternates": {"html": "https://wpnews.pro/news/profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through", "markdown": "https://wpnews.pro/news/profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through.md", "text": "https://wpnews.pro/news/profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through.txt", "jsonld": "https://wpnews.pro/news/profile-graph-memory-for-llm-agents-implicit-cross-entity-traversal-through.jsonld"}}