🧠 LastIn AI: The Fluff Whisperer

Academic paper analysis that separates the wheat from the chaff... and burns the chaff 🔥

🚀 Features

🤖 Fluff Annihilation Engine: Detects filler content with surgical precision (and sass)
📊 BS-o-Meter™: Rates papers on our patented How-Dare-You-Waste-My-Time™ scale
🔍 Context-Aware Roasting: Reads between LaTeX equations to ask "where's the beef?"
🛠 Self-Healing Storage: Fixes databases while judging your citation choices
🕰 Time-Aware Snark: Remembers when papers promised "revolutionary results" (spoiler: they didn't)

⚙️ Installation

# Clone repository (contains 0% filler content)
git clone https://github.com/your-lab/lastin-ai.git
cd lastin-ai

# Install dependencies (virtual env recommended)
pip install -r requirements.txt  # Includes anti-fluff field generators

# Initialize the system (PostgreSQL required)
python -m src.main init-db  # Creates tables and your first existential crisis

🔧 Configuration

Rename .env.example to .env and feed it:

OPENAI_API_KEY=sk-your-key-here  # We solemnly swear to misuse this responsibly
DEEPSEEK_API_KEY=sk-moon-shot     # For when regular snark isn't enough

DB_HOST=localhost                # Where we store academic sins
DB_NAME=paper_analysis           # Schema designed during a heated Twitter debate

🧠 Usage

# Harvest papers like a combine harvesting buzzwords
python -m src.main fetch --categories cs.AI --days 7

# Query your collection with sass
python -m src.main query "papers using 'transformative' unironically"

Pro Tip: Add --loglevel DEBUG to watch the AI question authors' life choices in real-time.

🏗 Architecture

lastin-ai/
├── src/
│   ├── data_acquisition/   # Where PDFs go to get judged
│   ├── storage/            # Home of our disappointment vector space
│   ├── utils/              # AgentController lives here (the Simon Cowell of AIs)
│   └── main.py             # The "I Regret Nothing" button

Our Agent Controller handles:

🧨 Fluff detection with precision-guided snark
📉 Quality metrics that hurt authors' feelings
🧮 Vector math that judges papers by their covers

🌌 Roadmap

Automated "Why Was This Published?" report generator
Sarcasm intensity slider (default: academic)
Support for papers that cite "inspired by ChatGPT" as methodology
Integration with Imposter Syndrome Detection API

🌟 Acknowledgments

ArXiv for the raw material
GPT-4 for pretending our prompts make sense
The letter F - for being the first character in "Fluff"

Disclaimer: Our fluff scores may correlate with your citation count. Not liable for sudden drops in academic self-esteem. 📉😭
P.S. Coffee-free since 2023 - runs on pure snark 🤖☕️➡️🗑️
P.P.S. Yes, we know the irony of an AI judging human creativity 🤖🔄🧑🎓

3.1 KiB Raw Blame History Unescape Escape