Ruwan Wickramarachchi

Graph-as-Policy (GaP) is a new variant of Agentic Robotics from @NVIDIA and @UCBerkeley that builds computation graphs (like ROS) to ensure modularity, manage complexity, and facilitate interpretability. 🧵Open code and paper: graph-robots.github.io/gap/ #Robotics #Automation #AgenticAI #EmbodiedAI

Very happy to be part of this effort. Graph-as-Policy (GaP), an agentic harness for robot learning, introduces the Variational Automation (VA) task class and eight open VA benchmarks, with a focus on complex industrial automation tasks.

🎉🎊 Our work 📜 Exploring the Abilities of Large Language Models to Solve Proportional Analogies via Knowledge-Enhanced Prompting accepted @coling2025 ! W/ @IwThilini @ruwantw @vnnm404 @VinijaJain @i_amanchadha @amitava_santu @amit_p #Coling2025 📜: arxiv.org/abs/2412.00869 🧵

AI Podcast: rss.com/podcasts/ai-pa…

Check out our latest article introducing DSceneKG, a suite of knowledge graphs built from real-world, multi-modal driving scene data, and its applications across multiple Neurosymbolic AI tasks. #NeurosymbolicAI #KnowledgeGraphs #DrivingScenes #AutonomousDriving

A comprehensive survey on rare event prediction aihub.org/2023/11/22/a-c…

Put together a list for folks like me who are interested in Neurosymbolic (neural net + knowledge representation/logic) AI and broader reasoning capabilities than the current crop of LLMs and stable diffusion models seem to evidenc…lnkd.in/gdcMfe-s lnkd.in/ggtTw8Sb

Our upcoming #ACL2023 Findings paper on "ANALOGICAL - A New Benchmark for Analogy of Long Text for Large Language Models": arxiv.org/abs/2305.05050 #ACL2023 #NLProc #Analogy

Join my group to work on compelling real-world problems. Yes, not late to apply! I am accepting PhD students w/ curious minds and an interest in understanding human behavior, building models w/ ML/DL, NLP, and KG. Drop me a DM if interested! Please spread the word! 1/2 👇

@GaryMarcus @tdietterich @billyperrigo We have created an open source dataset on toxic behavior in the USA under approved IRB without using any sweatshop or subjecting humans to psychologically punishing tasks: see the #1 publication at: wiki.aiisc.ai/index.php/Cont…

I gave ChatGPT questions from my Knowledge Graph class midterm. It always seemed to understand them and gave confident answers with well-formed explanations, but about half were wrong. Still impressive considering it never came to class. #ChatGPT #AI #LLM #NLP #knowledgeGraph

Check out our recent paper on the proposition of using knowledge graphs and entity prediction to improve machine perception in autonomous systems. #NeurosymbolicAI #AutonomousDriving #Industry40 #KEP

Interested in exploring knowledge-based entity prediction to improve machine perception in autonomous systems? Check out the article “Knowledge-Based Entity Prediction for Improved Machine Perception in Autonomous Systems” at lnkd.in/gHZzhF-d

The tutorial program can be found here: kl4ad.github.io/2022/

Interested in using knowledge graphs, neuro-symbolic AI, and their applications within autonomous driving? Consider attending our tutorial on "Knowledge-infused Learning for Autonomous Driving (#KL4AD)" at #ISWC2022 on Monday, 24th October, from 08:00 - 11:30 EST.

Seeing the future collide with the present is so fun.

Knowledge-based Entity Prediction for Improved Machine Perception in Autonomous Systems deepai.org/publication/kn… by @ruwantw et al. #MachineLearning #ComputerVision

Our recent paper on #KnowledgeInfusedLearning for entity prediction in #autonomousdriving scenes.

Exciting talk by @coryhenson on Knowledge-infused Learning and Autonomous Vehicles: #KiL2021 #KGC2021 #AIIsC aiisc.ai/KiL2021/speake…

We are #hiring. Know anyone who might be interested? Background: aiisc.ai lnkd.in/ecBsk72 Details, and where to apply: lnkd.in/d7iB3sY [ advanced analysis of sensor data using deep learning methods and synthetic control; n…lnkd.in/dd-9Cky