From Natural Language to Executable Narsese: A Neuro-Symbolic Benchmark and Pipeline for Reasoning with NARS
M. Gabriel, P. Wang · arXiv preprint
new
01
About
I'm a Data Scientist and AI/ML engineer currently serving as Senior Applications Developer for the Commonwealth of Pennsylvania, where I lead technical delivery for CWDS / PA CareerLink — an enterprise workforce platform supporting multiple state agencies.
In parallel, I'm completing my Ph.D. at Temple University with Prof. Pei Wang, researching neuro-symbolic reasoning, retrieval-augmented generation, and metacognitive confidence in large language models. I've previously contributed applied AI research at Cisco and NASA / JPL, and taught computer science at Harrisburg University from 2014 through 2026.
9
Publications
21
Citations
11+
Years Teaching
5
Research Orgs
02
Research Interests
Neuro-Symbolic AI & NARS
Combining the Non-Axiomatic Reasoning System with deep neural networks for explainable, resource-bounded inference under uncertainty.
Retrieval-Augmented Generation (RAG)
Adaptive RAG with metacognitive controllers — separating capability, stability, and entity-familiarity signals to decide when to retrieve.
LLM Confidence & Uncertainty
Self-consistency, aleatoric and epistemic uncertainty estimation grounded in Bandura's self-efficacy theory and NARS truth values.
Language Grounding & Semantic Gap
Why language grounding remains unsolved between human and artificial agents — and neuro-symbolic pipelines that translate natural language to executable Narsese.
Computer Vision & Perception
Real-time object detection, multi-object tracking, and sensor fusion (camera / RADAR / GPS) for autonomous and assistive systems.
03
Selected Publications
The Semantic Gap Between Human and Artificial Agents: Why Language Grounding Remains Unsolved
M. Gabriel
—
"Winning could mean success, yet losing doesn't mean failure" — Using a mobile serious game to facilitate science learning in middle school
S. Liu, B. Grey, M. Gabriel · Frontiers in Education 8, 1164462
4 cites
NARS in TruePal: a trusted and explainable AGI partner for first responders
C. Hahm, M. Gabriel, P. Hammer, P. Isaev, P. Wang · Tech. Report 19, Temple AGI Team
2 cites
Automated spatiotemporal modeling for real-time data-driven actionable insights
H. Latapie, M. Gabriel, S. Srinivasan, R. Kompella, K. R. Thórisson, P. Wang · SAI Intelligent Systems Conference
1 cite
Hybrid AI for IoT Actionable Insights & Real-Time Data-Driven Networks
H. Latapie, M. Gabriel, R. Kompella · Int'l Workshop on Self-Supervised Learning
7 cites
Wildlife Detection and Recognition in Digital Images Using YOLOv3
M. Gabriel, S. Cha, N. Y. B. Al-Nakash, D. Yun · IEEE Cloud Summit
7 cites
21 citations · h-index 3 · 9 articles
View on Google Scholar
04
Experience
Senior Applications Developer
Commonwealth of Pennsylvania — Harrisburg, PA
Jan 2026 — Present
Lecturer, Computer Science
Harrisburg University of Science & Technology
Aug 2014 — Jan 2026
Software Engineer III · Research & Engineering
Cisco
Oct 2022 — May 2023
Research Assistant & Software Engineer
NASA · JPL & Caltech collaboration
Oct 2020 — Sep 2021
Research Assistant, Artificial Intelligence
Temple University — with Dr. Pei Wang
Jan 2021 — Dec 2021
Software Engineer (Full-Stack)
The APAK Group — Hershey Center for Applied Research
Jan 2012 — Feb 2015
05
Tech Stack
Languages
AI / ML
Data
Cloud / DevOps
Backend
Research
06
Teaching
Artificial Intelligence
Search, planning, ML fundamentals
Machine Learning
Models, training, evaluation
Computer Vision
OpenCV, deep vision, tracking
Mobile Development
Android, Kotlin, Jetpack
Advanced Databases
SQL/NoSQL, modeling, indexing
Algorithms & Data Structures
Complexity, problem solving
07
Education
Ph.D. in Computer & Information Sciences (AI)
Temple University — advised by Prof. Pei Wang
M.S., Information Systems Engineering & Management
Harrisburg University of Science & Technology
B.S., Computer Science
Arab Academy for Science, Technology & Maritime Transport
08
Get in touch
Open to research collaborations, technical conversations, and consulting. Best reached by email — currently based in Mechanicsburg, PA.