Member proposals
What the collective is building next.
Proposal outline · members only
1 · Problem: what and why, 2–4 sentences
2 · Approach: model / method, and why this one
3 · Dataset: source, size, preprocessing
4 · Metrics: how you'll know it worked
5 · Compute: training time, memory, dependencies
6 · Deliverables: model, paper, demo
7 · Risk: what breaks first, and plan B
- P-01SE(3) airfoil diffusionPDF ↗
- P-02SD-Former · biosignalsPDF ↗
- P-03EPGNN · seismic graph networkPDF ↗
- P-04Hybrid weather forecastingPDF ↗
- P-05Earthquake predictorPDF ↗
- P-06CC-FNO-E · fluid dynamicsPDF ↗
- P-07VAE-BO drag reductionPDF ↗
- P-08HF-Bench · heart failurePDF ↗
- P-09Species extinction riskPDF ↗
- P-10Data-mixture interpolationPDF ↗
- P-11TriHet-PyPDF ↗
- P-12SympScalePDF ↗
- P-13MnemosynePDF ↗
- P-14Ankur SharmaPDF ↗
- P-15Hybrid extreme-weather forecastingPDF ↗
- P-16Arya AddagarlaPDF ↗
- P-17OliviaPDF ↗
- P-18Combined proposals · 33 projectsPDF ↗
- P-19DTLMUW waveform discoveryPDF ↗
- P-20FlutterFormPDF ↗
- P-21HF-Bench · DL vs. classical MLPDF ↗
- P-22Pick Two · clinical triagePDF ↗
- P-23SOFA UnpackedPDF ↗
- P-24Observable but Not SteerablePDF ↗
- P-25J-space global-workspace auditPDF ↗
- P-26REFLEX-RLVRPDF ↗
- P-27InterveneFMPDF ↗
- P-28Conflict · ablation-choice auditPDF ↗
- P-29Cell2CircuitPDF ↗
- P-30ITA · cross-species posePDF ↗
- P-31GrokScalePDF ↗
- P-32Unidentifiability oraclePDF ↗
- P-33Certified LoRA merge screeningPDF ↗
- P-34tensor-net-lmPDF ↗