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

Files changed (2) hide show
  1. app.py +4 -4
  2. 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="gaia-benchmark"
21
- DATA_DATASET = f"{OWNER}/GAIA"
22
- INTERNAL_DATA_DATASET = f"{OWNER}/GAIA_internal"
23
- SUBMISSION_DATASET = f"{OWNER}/submissions_internal"
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">GAIA Leaderboard</h1>"""
2
-
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
@@ -21,9 +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
- CITATION_BUTTON_LABEL = "Copy the following snippet to cite these results"
26
- CITATION_BUTTON_TEXT = r"""@misc{mialon2023gaia,
 
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},