Clarify the question first
Write down the context, assumptions, and objective before choosing tools.
FIELD NOTE · 01
Maolin Sun
Mendulin / Smiling
Mechanics & numerical simulation · Learning deep RL
My background spans geotechnical research and offshore-wind foundation engineering. I am now moving through machine learning, deep learning, and reinforcement learning toward neuromusculoskeletal modeling and control; my current focus is CS285.
NOW
Studying CS285
Working through core deep-RL methods, derivations, and implementations. The MS-Human-700 reproduction has not started.
SELECTED NOTES
Research and learning notes
Only completed research and learning work that is genuinely underway.
A modified strain wedge model for plastic development of reinforced concrete piles in sand
A modified strain-wedge model that incorporates plastic development in reinforced-concrete piles and is validated against tests and finite-element analysis.
From Ground to Muscle: The Start of a Learning Transition
Why I am moving from geotechnical engineering toward neuromusculoskeletal modeling, and how I plan to document a transition still in progress.
WORKING METHOD
How I work
Keep the reasoning visible
Record model choices, parameter sources, and failed paths so results can be checked.
Validate through small projects
Turn learning into experiments that can run, compare, and improve.