9 days ago
-
Add Gate anchor-calibration audit, split-half validation and geometric sensitivity
-
Add 800 audited seed and input-format robustness outputs with paired evaluation
-
Update historical alignment audit to reflect completed original-image scoring
-
Add completed metric ablations, paired statistics, anchor calibration and source audits
-
Add GPT real-photo B/C outputs and four-condition evaluation
10 days ago
-
Add five real-photo Track A baselines and matched evaluation
-
Add authorized rebuttal P0 conditioning controls and material-transfer baselines
-
Clarify fixed snapshot scope and pending final evaluation
-
Back up in-progress experiment outputs (1050/1600 images)
-
Back up in-progress experiment outputs (1000/1600 images)
-
Back up in-progress experiment outputs (900/1600 images)
-
Back up in-progress experiment outputs (800/1600 images)
-
Back up in-progress experiment outputs (700/1600 images)
-
Back up in-progress experiment outputs (600/1600 images)
-
Back up in-progress experiment outputs (500/1600 images)
-
Back up in-progress experiment outputs (400/1600 images)
-
Back up in-progress experiment outputs (300/1600 images)
-
Back up in-progress experiment outputs (200/1600 images)
-
Back up in-progress experiment outputs (100/1600 images)
-
Add three open-model real_photo C+ runs and paired five-model evaluation
-
Add paired real_photo C+ evaluation for GPT-Image-2 and Nano-Banana
-
Add 100 GPT-Image-2 C+ outputs for real_photo
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
-
Upload dataset
