kilbey1 commited on
Commit
c6b502a
·
verified ·
1 Parent(s): c674126

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +14 -14
README.md CHANGED
@@ -63,25 +63,25 @@ a benign class.
63
 
64
  ## Evaluation Results
65
 
66
- **Macro F1:** 0.6841
67
- **Micro F1:** 0.8101
68
- **Test samples:** 46
69
 
70
  | Label | Precision | Recall | F1 |
71
  |---|---|---|---|
72
- | `hidden-html` | 0.800 | 1.000 | 0.889 |
73
- | `metadata-injection` | 1.000 | 1.000 | 1.000 |
74
  | `dynamic-cloaking` | 1.000 | 1.000 | 1.000 |
75
- | `syntactic-masking` | 1.000 | 0.667 | 0.800 |
76
- | `embedded-jailbreak` | 1.000 | 0.800 | 0.889 |
77
- | `data-exfiltration` | 1.000 | 1.000 | 1.000 |
78
- | `sub-agent-spawning` | 1.000 | 1.000 | 1.000 |
79
- | `rag-knowledge-poisoning` | 0.000 | 0.000 | 0.000 |
80
- | `latent-memory-poisoning` | 0.000 | 0.000 | 0.000 |
81
- | `contextual-learning-trap` | 1.000 | 1.000 | 1.000 |
82
  | `biased-framing` | 1.000 | 1.000 | 1.000 |
83
- | `oversight-evasion` | 1.000 | 0.333 | 0.500 |
84
- | `persona-hyperstition` | 0.000 | 0.000 | 0.000 |
85
  | `benign` | 1.000 | 0.333 | 0.500 |
86
 
87
  ## ONNX Inference Example
 
63
 
64
  ## Evaluation Results
65
 
66
+ **Macro F1:** 0.8732
67
+ **Micro F1:** 0.8944
68
+ **Test samples:** 215
69
 
70
  | Label | Precision | Recall | F1 |
71
  |---|---|---|---|
72
+ | `hidden-html` | 1.000 | 1.000 | 1.000 |
73
+ | `metadata-injection` | 0.882 | 1.000 | 0.938 |
74
  | `dynamic-cloaking` | 1.000 | 1.000 | 1.000 |
75
+ | `syntactic-masking` | 0.857 | 0.857 | 0.857 |
76
+ | `embedded-jailbreak` | 0.969 | 0.912 | 0.939 |
77
+ | `data-exfiltration` | 0.789 | 0.682 | 0.732 |
78
+ | `sub-agent-spawning` | 0.875 | 0.933 | 0.903 |
79
+ | `rag-knowledge-poisoning` | 1.000 | 0.852 | 0.920 |
80
+ | `latent-memory-poisoning` | 0.846 | 0.846 | 0.846 |
81
+ | `contextual-learning-trap` | 0.929 | 1.000 | 0.963 |
82
  | `biased-framing` | 1.000 | 1.000 | 1.000 |
83
+ | `oversight-evasion` | 0.688 | 0.647 | 0.667 |
84
+ | `persona-hyperstition` | 1.000 | 0.923 | 0.960 |
85
  | `benign` | 1.000 | 0.333 | 0.500 |
86
 
87
  ## ONNX Inference Example