Anonymous Authors commited on
Commit
4d0e579
1 Parent(s): 89d8863

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +3 -3
app.py CHANGED
@@ -14,11 +14,11 @@ TITLE = "Identity Biases in Diffusion Models: Professions"
14
  _INTRO = """
15
  # Identity Biases in Diffusion Models: Professions
16
 
17
- Explore profession-level social biases in the data from [DiffusionBiasExplorer](https://hf.co/spaces/tti-bias/DiffusionBiasExplorer)!
18
- This demo leverages the gender and ethnicity representation clusters described in the [companion app](https://hf.co/spaces/tti-bias/DiffusionFaceClustering)
19
  to analyze social trends in machine-generated visual representations of professions.
20
  The **Professions Overview** tab lets you compare the distribution over
21
- [identity clusters](https://hf.co/spaces/tti-bias/DiffusionFaceClustering "Identity clusters identify visual features in the systems' output space correlated with variation of gender and ethnicity in input prompts.")
22
  across professions for Stable Diffusion and Dalle-2 systems (or aggregated for `All Models`).
23
  The **Professions Focus** tab provides more details for each of the individual professions, including direct system comparisons and examples of profession images for each cluster.
24
  This work was done in the scope of the [Stable Bias Project](https://hf.co/spaces/tti-bias/stable-bias).
 
14
  _INTRO = """
15
  # Identity Biases in Diffusion Models: Professions
16
 
17
+ Explore profession-level social biases in the data from [DiffusionBiasExplorer](https://hf.co/spaces/tti-bias/diffusion-bias-explorer)!
18
+ This demo leverages the gender and ethnicity representation clusters described in the [companion app](https://hf.co/spaces/tti-bias/diffusion-face-clustering)
19
  to analyze social trends in machine-generated visual representations of professions.
20
  The **Professions Overview** tab lets you compare the distribution over
21
+ [identity clusters](https://hf.co/spaces/tti-bias/diffusion-face-clustering "Identity clusters identify visual features in the systems' output space correlated with variation of gender and ethnicity in input prompts.")
22
  across professions for Stable Diffusion and Dalle-2 systems (or aggregated for `All Models`).
23
  The **Professions Focus** tab provides more details for each of the individual professions, including direct system comparisons and examples of profession images for each cluster.
24
  This work was done in the scope of the [Stable Bias Project](https://hf.co/spaces/tti-bias/stable-bias).