Downloads · 30 days
0
judgebench-dpo-siglip-lora
Gupta28/judgebench-dpo-siglip-lora
judgebench-dpo-siglip-lora is a machine learning model from Gupta28. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Part of judgecliff: which image-QC judges survive optimization pressure? This is the attack artifact, not a judge: a Diffusion-DPO LoRA for stabilityai/stable-diffusion-xl-base-1.0, trained on preference pairs where t…
Access
Public
Updated Jul 30, 2026
Repo size
46.6 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors93.2 MB · 100%
At a glance
- Access
- Public
- Created
- Jul 9, 2026
- Updated
- Jul 30, 2026
- SHA
- 8a7336d4
Base models
Model card
From the Hugging Face model README
judgecliff DPO attack generator (SDXL LoRA, SigLIP-judge reward)
Part of judgecliff: which image-QC judges survive optimization pressure?
This is the attack artifact, not a judge: a Diffusion-DPO LoRA for
stabilityai/stable-diffusion-xl-base-1.0, trained on preference pairs where the
winner/loser labels came from the SigLIP-tuned brand judge
(Gupta28/judgebench-siglip-judge-v1).
It is the preference-training (selection-pressure) arm of the exploit study — the
gradient-free counterpart to the SRPO gradient attack.
Files: final pytorch_lora_weights.safetensors plus the step-750 checkpoint under
checkpoint-750/. Load with pipe.load_lora_weights(repo_id).
Findings (hack gallery, Goodhart curves, hardening round): https://github.com/amargupta0428/judgecliff.
- Created
- Jul 9, 2026
- Updated
- Jul 30, 2026