LabVIEW Autocomplete
summary
Converts LabVIEW block diagrams into directed graphs, counts adjacent block pairs, and builds an order-1 Markov chain that suggests the most frequent next node; live demo shows predictions in milliseconds and auto‑wiring via Quick Drop. Model is tiny (few megabytes), runs entirely on your device, supports custom libraries by reindexing examples, and is open source. Training is additive but can take hours for a full corpus; practical but less context‑aware than transformer systems.
AI-written summary · about
chapters (13)
- 0:00LabVIEW Autocomplete intro
- 1:41What is LabVIEW Autocomplete
- 5:13Markov chains for prediction
- 6:58Converting VIs to state sequences
- 9:02Building prediction model
- 10:52Live autocomplete demo
- 13:24Testing with custom libraries
- 14:54Structures and data flow
- 17:25Predictions from data flow
- 18:31Real-world usefulness
- 20:04Advantages and limitations
- 21:52Future possibilities
- 23:41Q&A
related talks
08m4x7 G AI Assistant for LabVIEW (Project Code Name)
51mTest Workflows and Generative AI
56mMy AI Learning Journey
46mCode + AI: Will Robots Take Our Coding Jobs? ML Applied to Programming
43mMachine Learning for Developer Productivity
22mUsing Machine Learning to Improve UI/UX
01h 01mStructural Language Models of Code
12mNeuroSync: Intent-Aware Code-Based Problem Solving via Direct LLM Understanding Modification