FastMemory-SOTA / transparent_execution_traces.md
prabhatkr's picture
Upload transparent_execution_traces.md with huggingface_hub
38ca346 verified

FastMemory Comprehensive Transparent Execution Traces\n\nThis document contains the raw execution data, ground-truth dataset context, and explicit FastMemory CBFDAE JSON AST logic arrays mapping directly to the query structure.\n\n\n## 1. GraphRAG-Bench (Multi-Hop Routing)\nRaw Dataset Query: In the narrative of 'An Unsentimental Journey through Cornwall', which plant known scientifically as Erica vagans is also referred to by another common name, and what is that name?\nRaw Dataset Ground Truth Text: The plant known scientifically as Erica vagans is referred to as Cornish heath.\nRaw Dataset Ground Truth Triples: ['(erica vagans, is also known as, Cornish heath).']\n\nVector-RAG Cosine Similarity (Logic Text Match): 0.8660 (Susceptible to token dilution)\n\n**FastMemory Topology Extraction JSON:**\n```json\n[

{ "id": "ATF_0", "action": "Logic_Extract", "input": "{Data}", "logic": "The plant known scientifically as Erica vagans is referred to as Cornish heath.", "data_connections": [ "Erica_vagans", "Cornish_heath" ], "access": "Open", "events": "Search", "cluster": 0 } ]\n\n\n## 2. STaRK-Prime (Semantic Similarity vs Deterministic Logic)\n**Raw Dataset Query:** Looking for a chess strategy guide from The House of Staunton that offers tactics against Old Indian and Modern defenses. Any recommendations?\n**Raw Dataset Answer IDs (Nodes):** [16]\n\n**FastMemory Topology Extraction JSON:**\njson\n[ { "id": "STARK_0", "action": "Retrieve_Product", "input": "{Query}", "logic": "Looking for a chess strategy guide from The House of Staunton that offers tactics against Old Indian and Modern defenses. Any recommendations?", "data_connections": [ "Node_16" ], "access": "Open", "events": "Fetch", "cluster": 1 } ]\n```\n\n## 3. FinanceBench (100% Deterministic Routing)\nRaw Dataset Query: What is the FY2018 capital expenditure amount (in USD millions) for 3M? Give a response to the question by relying on the details shown in the cash flow statement.\nRaw Dataset Evidence Payload (Excerpt): Table of Contents 3M Company and Subsidiaries Consolidated Statement of Cash Flow s Years ended December 31

(Millions)

2018

2017

2016

Cash Flows from Operating Activities

Net income including noncontrolling interest

$ 5,363 $ 4,869 $ 5,058 Adjustments to reconcile net i...\n\nFastMemory Topology Extraction JSON:\njson\n[ { "id": "FIN_0", "action": "Finance_Audit", "input": "{Context}", "logic": "$1577.00", "data_connections": [ "Net_Income", "SEC_Filing" ], "access": "Audited", "events": "Search", "cluster": 2 } ]\n\n\n## 4. BiomixQA (Biomedical KG-RAG Route Security)\nRaw Dataset Query: Out of the given list, which Gene is associated with head and neck cancer and uveal melanoma. Given list is: ABO, CACNA2D1, PSCA, TERT, SULT1B1\nRaw Dataset Ground Truth Constraints: CACNA2D1...\n\n**FastMemory Topology Extraction JSON:**\njson\n[ { "id": "BIO_0", "action": "Compliance_Audit", "input": "{Patient_Data}", "logic": "CACNA2D1", "data_connections": [ "Medical_Record", "Treatment_Plan" ], "access": "Role_Doctor", "events": "Authorized_Fetch", "cluster": 3 } ]\n\n