rohhaiil commited on
Commit
3d62f55
·
verified ·
1 Parent(s): a3b9ffd

Upload dataset with hf_xet

Browse files
Files changed (1) hide show
  1. README.md +0 -12
README.md CHANGED
@@ -28,30 +28,19 @@ Given SysML v2 code and relevant context (compiler error, or relevant domain rul
28
  ## Fields
29
 
30
  - `id`: Unique identifier for each dataset instance.
31
-
32
  - `source_id`: Identifier of the original (clean) code example from which this instance was derived.
33
-
34
  - `mutation_category`: High-level category of the applied mutation:
35
  - `domain`: domain-specific semantic changes
36
  - `syntax`: syntactic errors
37
  - `none`: no mutation applied
38
-
39
  - `mutation_type`: Specific mutation operator used to generate the erroneous code.
40
-
41
  - `bad_code`: SysML v2 code potentially containing injected errors.
42
-
43
  - `good_code`: Correct version of the code.
44
-
45
  - `diff_patch`: Unified diff representing the transformation from `bad_code` to `good_code`.
46
-
47
  - `base_prompt`: Base prompt template used for fine-tuning; contains `bad_code` and compiler error in case of `syntax` mutations.
48
-
49
  - `prompt`: Prompt containing additional context (e.g., relevant domain rules), for `domain` and `none` mutations, where no compiler error occurs.
50
-
51
  - `code_response`: Repair in full code form.
52
-
53
  - `patch_response`: Repair in diff/patch format.
54
-
55
  - `length`: Total number of tokens in the full training sequence (prompt + response).
56
 
57
  ## Data Creation
@@ -74,7 +63,6 @@ This dataset is released under CC BY 4.0. Attribution to the original authors is
74
  ## Citation
75
 
76
  GitHub Repository: [SysMLv2 Repair with KG-SLMs](https://github.com/rohailamalik/SysMLv2-repair-with-KG-SLMs)
77
-
78
  ```bibtex
79
  @inproceedings{alshami2026sysml,
80
  title={Automated Semantic Fault Localization in SysML v2: A Human-in-the-Loop Framework Using Knowledge-Graph Augmented LLMs},
 
28
  ## Fields
29
 
30
  - `id`: Unique identifier for each dataset instance.
 
31
  - `source_id`: Identifier of the original (clean) code example from which this instance was derived.
 
32
  - `mutation_category`: High-level category of the applied mutation:
33
  - `domain`: domain-specific semantic changes
34
  - `syntax`: syntactic errors
35
  - `none`: no mutation applied
 
36
  - `mutation_type`: Specific mutation operator used to generate the erroneous code.
 
37
  - `bad_code`: SysML v2 code potentially containing injected errors.
 
38
  - `good_code`: Correct version of the code.
 
39
  - `diff_patch`: Unified diff representing the transformation from `bad_code` to `good_code`.
 
40
  - `base_prompt`: Base prompt template used for fine-tuning; contains `bad_code` and compiler error in case of `syntax` mutations.
 
41
  - `prompt`: Prompt containing additional context (e.g., relevant domain rules), for `domain` and `none` mutations, where no compiler error occurs.
 
42
  - `code_response`: Repair in full code form.
 
43
  - `patch_response`: Repair in diff/patch format.
 
44
  - `length`: Total number of tokens in the full training sequence (prompt + response).
45
 
46
  ## Data Creation
 
63
  ## Citation
64
 
65
  GitHub Repository: [SysMLv2 Repair with KG-SLMs](https://github.com/rohailamalik/SysMLv2-repair-with-KG-SLMs)
 
66
  ```bibtex
67
  @inproceedings{alshami2026sysml,
68
  title={Automated Semantic Fault Localization in SysML v2: A Human-in-the-Loop Framework Using Knowledge-Graph Augmented LLMs},