SabaPivot's picture
|
download
raw
2.42 kB
# Reproduction: What Makes a Strong Model?
Independent reproduction of ICML 2026 paper #23301 / arXiv:2606.01292 / OpenReview `ykWN4LG9vE`.
The authors did not link runnable code in the paper or arXiv source as of 2026-07-21. This bundle therefore implements the diagonal spectral mechanisms from the paper and a scaled real-architecture proxy independently.
## Scope
- `scripts/core_repro.py`: exact diagonal population dynamics for Theorems 1, 3, and 5, plus a Monte Carlo teacher-SGD / population-student experiment for weak-to-strong early stopping.
- `scripts/architecture_spectra.py`: scaled Figure 2(c) proxy using 1,000 images from `deedax/UTK-Face-Revised` and pretrained ResNet18, ResNet50, and ViT-B/16 features.
- `scripts/spectrum_fit_diagnostics.py`: log-log $R^2$ diagnostics for the persisted architecture spectra.
- `paper_source/`: official arXiv v1 source used to recover formulas and experiment details.
- `outputs/`: local smoke-test and downloaded Hugging Face Job results.
- `.trackio/logbook/`: canonical challenge logbook.
The full UTKFace training panels and ImageNet-based Figure 3(c,d) were not rerun: the paper does not release code/checkpoints, and those panels require several full pretrained feature sweeps and training runs. The architecture experiment here is explicitly a scaled proxy, not a full replication.
## Local rerun
```bash
uv venv .venv
UV_CACHE_DIR="$PWD/.uv-cache" uv pip install --python .venv/bin/python numpy
.venv/bin/python scripts/core_repro.py --output outputs/core --reps 64 --seed 23301
```
The real-architecture proxy is a PEP 723 script and can be run with:
```bash
uv run scripts/architecture_spectra.py --output outputs/architecture --samples 1000 --batch-size 32
```
## Hugging Face runs and artifacts
- Results bucket: https://huggingface.co/buckets/SabaPivot/repro-what-makes-a-strong-model-results
- Jobs canary: https://huggingface.co/jobs/SabaPivot/6a5f1ee413e6ef894d549757
- Core T4 run: https://huggingface.co/jobs/SabaPivot/6a5f2044d09dc1f57c6bed22
- Architecture T4 first attempt (audited output-directory failure): https://huggingface.co/jobs/SabaPivot/6a5f205113e6ef894d549760
- Architecture T4 successful retry: https://huggingface.co/jobs/SabaPivot/6a5f2198d09dc1f57c6bed3c
- UTKFace dataset: https://huggingface.co/datasets/deedax/UTK-Face-Revised
All repository-specific generated files stay under this reproduction directory.

Xet Storage Details

Size:
2.42 kB
·
Xet hash:
1d9f746ed308b1c7d2a62c6e0b0fce50ab5d05e425404d940878b31b29cc0e32

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.