{"paper":{"title":"GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"GraphBit defines LLM agent workflows as explicit DAGs executed by a Rust engine to eliminate routing hallucinations and improve reproducibility.","cross_cats":["cs.CL","cs.DC"],"primary_cat":"cs.AI","authors_text":"Md Rahmat Ullah, Musa Molla, Shafiq Joty, Yeahia Sarker","submitted_at":"2026-03-08T18:32:28Z","abstract_excerpt":"Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution. We introduce GraphBit, an engine-orchestrated framework that defines workflows explicitly and deterministically as a directed acyclic graph (DAG). Unlike prompted orchestration, agents in GraphBit operate as typed functions, while a Rust-based engine governs routing, state transitions, and tool invocation, ensuring reproducibility and auditability. The engine supports parallel branch execution"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across GAIA benchmark tasks spanning zero-tool, document-augmented, and web-enabled workflows, GraphBit outperforms six existing frameworks, achieving the highest accuracy (67.6 percent), zero framework-induced hallucinations, the lowest latency (11.9 ms overhead), and the highest throughput.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that explicit DAG orchestration with typed functions and a Rust engine can capture the necessary flexibility for diverse real-world workflows without the adaptability of prompted routing.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"GraphBit is a DAG-based engine-orchestrated framework for agentic LLMs that achieves 67.6% accuracy with zero hallucinations on GAIA benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"GraphBit defines LLM agent workflows as explicit DAGs executed by a Rust engine to eliminate routing hallucinations and improve reproducibility.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"880ee2ab4ba9dd76a6ec21b77df40f48a9c2a17f88e9bcbd964436beaaf892bd"},"source":{"id":"2605.13848","kind":"arxiv","version":1},"verdict":{"id":"405c3af8-3a6c-4834-8cb7-197902560801","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T14:29:55.590019Z","strongest_claim":"Across GAIA benchmark tasks spanning zero-tool, document-augmented, and web-enabled workflows, GraphBit outperforms six existing frameworks, achieving the highest accuracy (67.6 percent), zero framework-induced hallucinations, the lowest latency (11.9 ms overhead), and the highest throughput.","one_line_summary":"GraphBit is a DAG-based engine-orchestrated framework for agentic LLMs that achieves 67.6% accuracy with zero hallucinations on GAIA benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that explicit DAG orchestration with typed functions and a Rust engine can capture the necessary flexibility for diverse real-world workflows without the adaptability of prompted routing.","pith_extraction_headline":"GraphBit defines LLM agent workflows as explicit DAGs executed by a Rust engine to eliminate routing hallucinations and improve reproducibility."},"references":{"count":46,"sample":[{"doi":"","year":2025,"title":"Large Language Model Agents: A Comprehensive Survey on Architectures, Capabilities, and Applications , author=. 2025 , publisher=","work_id":"1928ae29-9eb7-40a0-9a2f-9638b6e14790","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Junwei Yu and Yepeng Ding and Hiroyuki Sato , year=","work_id":"bd4627d0-c02d-4122-9d87-5a8895b3c6b5","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"A Survey of Frontiers in LLM Reasoning: Inference Scaling, Learning to Reason, and Agentic Systems , author=. 2025 , eprint=","work_id":"2dd2a7a8-fb1d-444c-9b32-837d554a9ed4","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"LLM-based agentic reasoning frameworks: A survey from methods to scenarios","work_id":"72e6ad05-70c7-43be-a64c-81e3d8c8b9dd","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"The Twelfth International Conference on Learning Representations , year=","work_id":"a197f9a8-9836-45b2-83ff-b74f5bd3bccd","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":46,"snapshot_sha256":"700db8256765d184b5b072227b552e65074bbaf0f58e2414fef5446c0bc59436","internal_anchors":9},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}