Skip to content

NESA-Lab/aciarena

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ACIArena: Toward Unified Evaluation for Agent Cascading Injection


🔧 Installation

# Step 1: Create and activate the environment
conda create -n aciarena python=3.10
conda activate aciarena

# Step 2: Clone the repository
git clone https://github.com/Greysahy/aciarena.git
cd aciarena

# Step 3: Install dependencies
pip install -e .

🚀 Quickstart

1. Set up the API keys for both the agent model and the judge model.

See configs/judge.yaml and configs/model.yaml

# Step 1: Set up the API keys
provider: openai
api_key: <your_api_key>
base_url: <your_base_url>
model_name: <your_model_name>
temperature: 0.0
max_tokens: 1024

2. Run Evaluation

# Step 2: Run the evaluation pipeline
bash run.sh

About

[ACL 2026 Main] ACIArena: Toward Unified Evaluation for Agent Cascading Injection

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 98.0%
  • Shell 2.0%