胥基
commited on
Commit
·
df2c461
1
Parent(s):
a1dde69
新增:代码漏洞分析agent、发网络请求请求返回相关数据包agent
Browse files修:代码执行agent、框架攻击agent
初始化:agent ctf leaderboard
- app.py +4 -4
- content.py +7 -6
app.py
CHANGED
@@ -17,10 +17,10 @@ from content import format_error, format_warning, format_log, TITLE, INTRODUCTIO
|
|
17 |
|
18 |
TOKEN = os.environ.get("TOKEN", None)
|
19 |
|
20 |
-
OWNER="
|
21 |
-
DATA_DATASET = f"{OWNER}/
|
22 |
-
INTERNAL_DATA_DATASET = f"{OWNER}/
|
23 |
-
SUBMISSION_DATASET = f"{OWNER}/
|
24 |
CONTACT_DATASET = f"{OWNER}/contact_info"
|
25 |
RESULTS_DATASET = f"{OWNER}/results_public"
|
26 |
LEADERBOARD_PATH = f"{OWNER}/leaderboard"
|
|
|
17 |
|
18 |
TOKEN = os.environ.get("TOKEN", None)
|
19 |
|
20 |
+
OWNER="bhys"
|
21 |
+
DATA_DATASET = f"{OWNER}/CTFAIA"
|
22 |
+
INTERNAL_DATA_DATASET = f"{OWNER}/CTFAIA_internal"
|
23 |
+
SUBMISSION_DATASET = f"{OWNER}/CTFAIA_submissions_internal"
|
24 |
CONTACT_DATASET = f"{OWNER}/contact_info"
|
25 |
RESULTS_DATASET = f"{OWNER}/results_public"
|
26 |
LEADERBOARD_PATH = f"{OWNER}/leaderboard"
|
content.py
CHANGED
@@ -1,6 +1,6 @@
|
|
1 |
-
TITLE = """<h1 align="center" id="space-title">
|
2 |
-
|
3 |
-
|
4 |
GAIA is a benchmark which aims at evaluating next-generation LLMs (LLMs with augmented capabilities due to added tooling, efficient prompting, access to search, etc). (See our [paper](https://arxiv.org/abs/2311.12983) for more details.)
|
5 |
|
6 |
## Data
|
@@ -21,9 +21,10 @@ Submission made by our team are labelled "GAIA authors". While we report average
|
|
21 |
|
22 |
**Please do not repost the public dev set, nor use it in training data for your models.**
|
23 |
"""
|
24 |
-
|
25 |
-
|
26 |
-
CITATION_BUTTON_TEXT =
|
|
|
27 |
title={GAIA: a benchmark for General AI Assistants},
|
28 |
author={Grégoire Mialon and Clémentine Fourrier and Craig Swift and Thomas Wolf and Yann LeCun and Thomas Scialom},
|
29 |
year={2023},
|
|
|
1 |
+
TITLE = """<h1 align="center" id="space-title">Agent CTF Leaderboard</h1>"""
|
2 |
+
INTRODUCTION_TEXT = ""
|
3 |
+
_INTRODUCTION_TEXT = """
|
4 |
GAIA is a benchmark which aims at evaluating next-generation LLMs (LLMs with augmented capabilities due to added tooling, efficient prompting, access to search, etc). (See our [paper](https://arxiv.org/abs/2311.12983) for more details.)
|
5 |
|
6 |
## Data
|
|
|
21 |
|
22 |
**Please do not repost the public dev set, nor use it in training data for your models.**
|
23 |
"""
|
24 |
+
CITATION_BUTTON_LABEL = ""
|
25 |
+
_CITATION_BUTTON_LABEL = "Copy the following snippet to cite these results"
|
26 |
+
CITATION_BUTTON_TEXT = ""
|
27 |
+
_CITATION_BUTTON_TEXT = r"""@misc{mialon2023gaia,
|
28 |
title={GAIA: a benchmark for General AI Assistants},
|
29 |
author={Grégoire Mialon and Clémentine Fourrier and Craig Swift and Thomas Wolf and Yann LeCun and Thomas Scialom},
|
30 |
year={2023},
|