Spaces:

ProCreations
/

repro-finding-most-influential-sets

Running

App Files Files Community

repro-finding-most-influential-sets

4.95 MB

Ctrl+K

1 contributor

History: 35 commits

ProCreations

Replace the source-pinned claim-6 page with an independent 10-D Gaussian-mixture self-consuming-loop experiment

68e53cc verified 5 days ago

code
Claim 6: run the 10-D Gaussian-mixture self-consuming loop independently (was source-pinned only); alpha=0 degrades 9.4x vs 1.2x at alpha=1 over 80 generations 5 days ago
outputs
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
packaged_replay
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
pages
Replace the source-pinned claim-6 page with an independent 10-D Gaussian-mixture self-consuming-loop experiment 5 days ago
source
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
.gitattributes

2.05 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
BUNDLE_SHA256SUMS.txt

6.48 kB
Reconcile Hub LFS rules for diffusion model collapse reproduction 8 days ago
README.md

753 Bytes
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
SOURCE_PIN.txt

445 Bytes
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
bucket-icon.svg

413 Bytes
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago
build_manifest.py

952 Bytes
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
claims.json

1.41 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
index.html

1.87 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
logbook.css

29.6 kB
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago
logbook.js

77.3 kB
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago
logbook.json

5.33 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
official_claims.json

1.41 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
outputs_collapse_results.json

1.44 kB
Claim 6: run the 10-D Gaussian-mixture self-consuming loop independently (was source-pinned only); alpha=0 degrades 9.4x vs 1.2x at alpha=1 over 80 generations 5 days ago
poster_embed.html

1.65 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
reproduce.py

18.2 kB
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
requirements.txt

32 Bytes
Repurpose zero-loss duplicate for diffusion model collapse reproduction 8 days ago
trackio-logo-light.png

30 kB
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago
trackio-logo.png

55.6 kB
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago
trackio-wordmark-dark.png

89.8 kB
Publish independent ICML 2026 reproduction: Finding Most Influential Sets 12 days ago