alexbie98 commited on
Commit
d79e92b
·
verified ·
1 Parent(s): effe61d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -2
README.md CHANGED
@@ -8,11 +8,15 @@ pinned: false
8
  ---
9
  # ContinuousBench
10
 
11
- ContinuousBench measures progress in differentially private synthetic data.
 
 
 
 
12
 
13
  ContinuousBench has two tracks:
14
  * [Geminon](https://huggingface.co/datasets/ContinuousBench/Geminon): Fictional, Gemini-generated corpus
15
- * [News](https://huggingface.co/datasets/ContinuousBench/News): Scraped news articles from Septmember 2025
16
 
17
  Both datasets:
18
  * are designed to contain completely new information that models cannot answer
@@ -20,6 +24,7 @@ Both datasets:
20
 
21
 
22
  Generate a DP synthetic version of News or Geminon, then test it: https://github.com/plau666/ContinuousBenchEval.
 
23
  Our evaluation trains a model on your DP synthetic version, and then asks the paired QA to see if your DP synthetic data was capable of teaching a model the knowledge present in the original corpus.
24
 
25
 
 
8
  ---
9
  # ContinuousBench
10
 
11
+ [Blog post](https://peihanliu.com/posts/continuousbench.html) | [Arxiv](#)
12
+
13
+
14
+ **ContinuousBench** measures progress in differentially private synthetic data.
15
+
16
 
17
  ContinuousBench has two tracks:
18
  * [Geminon](https://huggingface.co/datasets/ContinuousBench/Geminon): Fictional, Gemini-generated corpus
19
+ * [News](https://huggingface.co/datasets/ContinuousBench/News): Scraped news articles from September 2025
20
 
21
  Both datasets:
22
  * are designed to contain completely new information that models cannot answer
 
24
 
25
 
26
  Generate a DP synthetic version of News or Geminon, then test it: https://github.com/plau666/ContinuousBenchEval.
27
+
28
  Our evaluation trains a model on your DP synthetic version, and then asks the paired QA to see if your DP synthetic data was capable of teaching a model the knowledge present in the original corpus.
29
 
30