SKM / RESEARCH
Questions worth pursuing.
Machine intelligence, representation, and the constraints that make research interesting.
5 entries
01
Software & Learning Systems / SELECTED
Evidence-Bounded Software Analysis
How semantic summaries and isolated evaluation can support inspectable code-change proposals.
Independent exploration
02Software & Learning Systems / SELECTED
Search and Learned Chess Evaluation
An experimental study of classical evaluation, neural accumulators, and phase-dependent blending.
Independent exploration
03Time-series Modeling / SELECTED
Self-Supervised EEG Representation Learning
An exploration of temporal contrastive learning, subject-level evaluation, and the limits of apparent gains.
Independent exploration
04Computer Vision
H-STAR
Sign language recognition: exploring how visual representations can capture the structure of signed communication.
Verification pending
05Efficient ML
MobileLoRA
Hardware-aware on-device adapter management under memory and compute constraints.
Verification pending