{"slug": "agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of", "title": "AgentKGV: Agentic LLM-RAG Framework with Two-Stage Training for the Fact Verification of Knowledge Graphs", "summary": "Researchers propose AgentKGV, an agentic LLM-RAG framework for verifying facts in knowledge graphs, using dynamic routing and iterative query rewriting. A two-stage training strategy combining distillation-based SFT and trajectory-level GRPO improves macro-F1 by 14.9 percentage points over single-turn RAG on the T-REx benchmark while reducing search calls by half.", "body_md": "arXiv:2607.09092v1 Announce Type: new\nAbstract: Knowledge graphs (KGs) are often automatically constructed from large-scale corpora, but they inevitably contain factual errors due to noisy sources and extraction failures, and verifying them reliably at industrial scale remains a critical challenge. To address this, we propose AgentKGV, the Agentic LLM-RAG framework for KG fact Verification, that integrates dynamic routing and iterative query rewriting, which handles surface-form mismatch in document-level retrieval. To make this framework more accurate and cost-efficient for industrial deployment, we further introduce a two-stage training strategy: turn-level distillation-based SFT that transfers reasoning ability from a large teacher model into a small model for stable query rewriting and reasoning, and trajectory-level GRPO that optimizes the search policy to reduce unnecessary retrieval at scale. On the long-tail-predicate split of the open-domain T-REx benchmark, our framework improves macro-F1 over single-turn RAG by 5.5 \\%p, and two-stage training does it further by 9.4 \\%p. GRPO also cuts the average number of search calls from 3.24 to 1.63 without lowering accuracy.", "url": "https://wpnews.pro/news/agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of", "canonical_source": "https://arxiv.org/abs/2607.09092", "published_at": "2026-07-13 04:00:00+00:00", "updated_at": "2026-07-13 04:17:05.267203+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "ai-agents"], "entities": ["AgentKGV", "T-REx"], "alternates": {"html": "https://wpnews.pro/news/agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of", "markdown": "https://wpnews.pro/news/agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of.md", "text": "https://wpnews.pro/news/agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of.txt", "jsonld": "https://wpnews.pro/news/agentkgv-agentic-llm-rag-framework-with-two-stage-training-for-the-fact-of.jsonld"}}