jayubba commited on
Commit
87f6e8a
路
verified 路
1 Parent(s): 8b4b108

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -14,7 +14,7 @@ It's live at [semblen.com](https://semblen.com), including an index of 2026 Sena
14
 
15
  Our technique models are fine-tuned DeBERTa-v3-large multi-label classifiers: 16 technique heads, plus one that estimates whether the charged language belongs to a quoted speaker rather than the outlet. They're the first stage of the pipeline; a language model then explains each flag in plain words.
16
 
17
- The models are private. Researchers can request access: email [jason@semblen.com](mailto:jason@semblen.com).
18
 
19
  ## How we work
20
 
@@ -22,4 +22,4 @@ The models are private. Researchers can request access: email [jason@semblen.com
22
  - We audit flags by hand. In our last blind audit, about 1 in 10 was wrong.
23
  - A flag is about how something is worded, not whether it's true or who's right.
24
 
25
- Semblen LLC 路 Austin, Texas 路 [jason@semblen.com](mailto:jason@semblen.com)
 
14
 
15
  Our technique models are fine-tuned DeBERTa-v3-large multi-label classifiers: 16 technique heads, plus one that estimates whether the charged language belongs to a quoted speaker rather than the outlet. They're the first stage of the pipeline; a language model then explains each flag in plain words.
16
 
17
+ The models are private. Researchers can request access: email [hello@semblen.com](mailto:hello@semblen.com).
18
 
19
  ## How we work
20
 
 
22
  - We audit flags by hand. In our last blind audit, about 1 in 10 was wrong.
23
  - A flag is about how something is worded, not whether it's true or who's right.
24
 
25
+ Semblen LLC 路 Austin, Texas 路 [hello@semblen.com](mailto:hello@semblen.com)