Interpreting the results
A successful API response will return a json record, like the one below, while an unsuccessful attempt will return an error code (see API specs for more details and examples).
{"score": 0.98,
"address": "0xE42CAD6FC883877A76A26A16ED92444AB177E306",
"createdAt": "2023-03-25T09:21:43Z",
"isMalicious": 0}
The json record includes four keys: address, createdAt, isMalicious, score.
score: the score returned is a trust score that lies in the range from 0 to 1; where a score of 1 is bad or malicious, and a score of 0 means no malicious patterns detected in the code. Any score between 0 and 1 represents a confidence value of the model. For example, the json record that the API returned above shows a score of 0.98, meaning that the model is 98% confident that the smart contract code is not malicious.
isMalicious: isMalicious is a flag for those who prefer a binary understanding. In this case, 0 means False (Not Malicious), and 1 means True (Malicious). To generate this flag, the model assigns True to every score that is greater than 0.5.
Last modified 5mo ago