Published artifacts

Architectures built from scratch.
Parameters counted by hand.

per-cell argmax · 54×96
01 / EDGE DETECTION

ANetV1 · per-region UAV detection at the edge

Avneh Bhatia

17,392 parameters, 150× smaller than YOLO134 img/s on an M-series GPU17 KB at int8, smaller than a JPEG
cl / α polar · re 1e5
02 / AEROSPACE

FoilCORE · in-context memory along airfoil polars

Avneh Singh Bhatia · Joshua Selvaraj

6,394 parametersbeats public NeuralFoil checkpointscausal in-AoA-order polar sequencing
pitch + heave · flutter onset
03 / AEROELASTICITY

FlutterForm · learned aeroelastic flutter prediction

Exea Labs

predicts flutter speed v_f and frequency ω_f directly<2.7% error on withheld sections50,000 aeroelastic training samples
learned directed graph · sparse
04 / BIOSIGNALS

SD-Former · sparse directed attention for physiological time series

Ankur Sharma

graph-conditioned biosignal architecturerobust under subject & device shiftcontinuous DAG learning as inductive bias
detection ∝ network topology
05 / SEISMOLOGY

EPGNN · multimodal seismic graph network for early signals

Arya Addagarla

detections propagate by topology, not thresholdslocal noise ≠ regional alertspeed and spatial precision

20+ member proposals in flight.

Browse proposals