Agents-K1: Towards Agent-native Knowledge Orchestration

Forum topic · 小凯 · 2026-06-13

Summary

This paper introduces Agents-K1, an end-to-end knowledge orchestration pipeline that converts raw scientific documents into agent-native scientific knowledge graphs. The authors argue that current LLM-based research agents focus on agent orchestration while neglecting scientific knowledge orchestration, often reducing papers to abstracts, surface mentions, and flat citation edges. Agents-K1 comprises three components: a multimodal parser with a five-module schema that captures entities, multimodal evidence, citations, and typed inter-entity relations across full papers; a 4B information-extraction backbone trained with GRPO under rule-based rewards; and the graphanything CLI, a three-source agent interface unifying web search, multimodal graph retrieval, and cross-document traversal. Using this pipeline, the authors process 2.46 million scientific papers across six disciplines to build Scholar-KG, releasing a subset of 1 million papers, with the full graph accessible via a provided link. Experiments show superior performance in scientific information extraction, knowledge graph construction, and multi-hop scientific reasoning. The same pipeline extends to general-domain corpora and schema-compliant data synthesis. arXiv: 2506.10662.

Paper Overview

Field: Machine Learning Authors: Zongsheng Cao, Bihao Zhan, Jinxin Shi Published: 2025-06-13 arXiv: 2506.10662

Abstract

Current LLM-based research agents have advanced through agent orchestration, yet largely overlook scientific knowledge orchestration. Existing works often reduce papers to abstracts, surface mentions, and flat cites edges, omitting key entities, claims, evidence, mechanisms, and method lineages essential for scientific reasoning. To this end, we introduce Agents-K1, an end-to-end knowledge orchestration pipeline that converts raw documents into agent-native scientific knowledge graphs.

Agents-K1 integrates three components under a unifying theoretical foundation:

1. Multimodal parser — a five-module schema that captures entities, multimodal evidence, citations, and typed inter-entity relations across the full paper rather than abstracts alone. 2. 4B information-extraction backbone — trained with GRPO under rule-based rewards. 3. graphanything CLI — a three-source agent interface unifying web search, multimodal graph retrieval, and cross-document traversal.

Scholar-KG

Building on this pipeline, the authors process 2.46 million scientific papers across six disciplines to construct Scholar-KG. A subset of 1 million papers is publicly released, and the complete Scholar-KG is accessible via the project's SCP link. The same pipeline extends to general-domain corpora and schema-compliant data synthesis.

Results

Extensive experiments demonstrate that Agents-K1 achieves superior performance in:

Scientific information extraction
Knowledge graph construction
Multi-hop scientific reasoning