Skip to content
View sunghunkwag's full-sized avatar
1 follower · 0 following

Achievements

Achievement: Pair Extraordinairex3Achievement: Quickdraw

Achievements

Achievement: Pair Extraordinairex3Achievement: Quickdraw

Block or report sunghunkwag

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sunghunkwag/README.md

Independent AI Researcher

I build and test alternative AI architectures — focused on compute-efficient design beyond transformer scaling.

Current work explores adaptive self-optimization and meta-reinforcement learning within real-world compute limits.

Contact: sunghunkwag@gmail.com

Pinned Loading

  1. ast-grammar-induction-prototype ast-grammar-induction-prototype Public

    A single-file recursive self-improvement engine that evolves Python programs through AST analysis and statistical grammar learning (EDA).

    Python 1

  2. structural-memory-field structural-memory-field Public

    Physics-based memory system using wound dynamics, crystallization, and erosion in continuous fields — no neural networks or learned parameters.

    Python

  3. DHC-SSM-Enhanced DHC-SSM-Enhanced Public

    Experimental state space model testbed probing linear-time (O(n)) assumptions for causal modeling beyond Transformers.

    Python 1

  4. SSM-MetaRL-TestCompute SSM-MetaRL-TestCompute Public

    Experimental meta-reinforcement learning framework exploring State Space Model (SSM) policies, test-time adaptation, and structured benchmarking.

    Python