Files changed (1) hide show
  1. README.md +22 -18
README.md CHANGED
@@ -3,29 +3,33 @@ license: mit
3
  configs:
4
  - config_name: train
5
  data_files:
6
- - split: train
7
- path: train.jsonl
8
  - config_name: single_test
9
  data_files:
10
- - split: train
11
- path: single_test.jsonl
12
  - config_name: multiple_test
13
  data_files:
14
- - split: train
15
- path: multiple_test.jsonl
16
-
17
- rai:
18
- - dataLimitations: The current benchmark focuses on code generation scenarios with verifiable outcomes. Extending it to broader domains, such as factual knowledge tasks, would help evaluate the generality of our findings in more realistic settings.
 
 
 
 
 
 
 
 
 
 
 
 
 
19
 
20
- - dataBiases: We do not identify significant systematic biases in the current benchmark construction.
21
-
22
- - personalSensitiveInformation: The dataset does not intentionally include personal or sensitive information.
23
-
24
- - dataUseCases: SE-Bench is intended for research on self-evolution and knowledge internalization.
25
-
26
- - dataSocialImpact: The benchmark provides a controlled and clean environment for studying the self-evolution capabilities of language models.
27
-
28
- - hasSyntheticData: true
29
  ---
30
 
31
  ## dataset structure
 
3
  configs:
4
  - config_name: train
5
  data_files:
6
+ - split: train
7
+ path: train.jsonl
8
  - config_name: single_test
9
  data_files:
10
+ - split: train
11
+ path: single_test.jsonl
12
  - config_name: multiple_test
13
  data_files:
14
+ - split: train
15
+ path: multiple_test.jsonl
16
+ rai:dataLimitations: >
17
+ The current benchmark focuses on code generation scenarios with verifiable
18
+ outcomes. Extending it to broader domains, such as factual knowledge tasks,
19
+ would help evaluate the generality of our findings in more realistic settings.
20
+ rai:dataBiases: >
21
+ We do not identify significant systematic biases in the current benchmark
22
+ construction.
23
+ rai:personalSensitiveInformation: |
24
+ The dataset does not intentionally include personal or sensitive information.
25
+ rai:dataUseCases: >
26
+ SE-Bench is intended for research on self-evolution and knowledge
27
+ internalization.
28
+ rai:dataSocialImpact: >
29
+ The benchmark provides a controlled and clean environment for studying the
30
+ self-evolution capabilities of language models.
31
+ rai:hasSyntheticData: true
32
 
 
 
 
 
 
 
 
 
 
33
  ---
34
 
35
  ## dataset structure