Search papers. Explore concepts. Visualize mathematics. Generate reviews.
Features • Quick Start • Agents • Visualizations • Web UI • Examples
"Can you explain Calabi-Yau compactification ┌─────────────┐
and find the latest papers on it?" │ Scholar │──→ arXiv Search
│ └──────┬──────┘
▼ │
┌─────────────┐ ┌─────────────┐ ┌──────▼──────┐
│ Theorist │────▶│ Synthesizer │────────────────▶│ Literature │
│ (LLM) │ │ (Review) │ │ Review │
└──────┬──────┘ └─────────────┘ └─────────────┘
│
▼
┌─────────────┐
│Mathematician│──→ 3D Calabi-Yau Plot
│ (Compute) │──→ Worldsheet Animation
└─────────────┘──→ AdS Space Visualization
StringTheoryAI is an AI research agent that helps physicists, students, and curious minds explore string theory. It combines:
- Real arXiv integration — search and analyze papers from hep-th, gr-qc, math-ph
- AI-powered explanations — deep, citation-backed answers from GPT-4o or Claude
- Mathematical visualizations — interactive 3D plots of Calabi-Yau manifolds, worldsheets, branes, and more
- Automated literature reviews — synthesize dozens of papers into structured reviews
Built for anyone from graduate students to active researchers.
| Feature | Description |
|---|---|
| Paper Search | Search arXiv across hep-th, gr-qc, math-ph with AI analysis |
| Concept Explorer | Ask anything about string theory — get rigorous, cited explanations |
| Concept Cards | Structured knowledge cards with equations, related topics, key papers |
| Compare Concepts | Side-by-side comparison of theories, dualities, approaches |
| Math Visualizations | Interactive 3D plots: Calabi-Yau, worldsheets, AdS, branes, tori |
| Literature Reviews | AI-generated research summaries from arXiv papers |
| Trend Analysis | Identify emerging directions across paper collections |
| Quiz Mode | Test your string theory knowledge at any level |
| Rich CLI | Beautiful terminal interface with tables and panels |
| Streamlit Dashboard | Full web UI with interactive plots |
pip install stringtheoryaiOr from source:
git clone https://github.com/gewenbo888/StringTheoryAI.git
cd StringTheoryAI
pip install -e ".[all]"cp .env.example .env
# Add your OpenAI or Anthropic API key# Search papers
stringtheory search "AdS/CFT black hole entropy"
# Explain a concept
stringtheory explain "What is T-duality in string theory?"
# Get a structured concept card
stringtheory explain "Calabi-Yau compactification" --concept-card
# Compare concepts
stringtheory compare "Type IIA" "Type IIB"
# Generate visualizations (no API key needed!)
stringtheory visualize calabi-yau
stringtheory visualize all
# Generate a literature review
stringtheory review "holographic entanglement entropy"
# Test your knowledge
stringtheory quiz "supersymmetry" --difficulty graduate
# Launch web dashboard
stringtheory uiSearches arXiv and provides AI-powered paper analysis.
from stringtheory.agents.scholar import Scholar
from stringtheory.config import ArxivConfig, LLMConfig
scholar = Scholar(LLMConfig(), ArxivConfig())
# Search
papers = await scholar.search("swampland conjecture")
# Analyze a paper
analysis = await scholar.analyze_paper(papers[0])
# Find related work
related = await scholar.find_related(papers[0])
# Search by author
witten_papers = await scholar.search_by_author("Edward Witten")Deep explanations with mathematical rigor.
from stringtheory.agents.theorist import Theorist
theorist = Theorist(LLMConfig())
# Detailed explanation
explanation = await theorist.explain("What is the AdS/CFT correspondence?")
# Structured concept card
concept = await theorist.get_concept("M-theory")
# Compare approaches
comparison = await theorist.compare("Loop quantum gravity", "String theory")
# Historical trace
history = await theorist.trace_history("The second superstring revolution")Mathematical computations and stunning 3D visualizations.
from stringtheory.agents.mathematician import Mathematician
math = Mathematician(LLMConfig(), VisualizationConfig())
# Generate visualizations
math.visualize_calabi_yau(n=5) # Calabi-Yau manifold
math.visualize_worldsheet(modes=4) # String worldsheet
math.visualize_torus() # T² compactification
math.visualize_ads_space() # Anti-de Sitter space
math.visualize_branes(n_branes=7) # D-brane stack
math.visualize_extra_dimensions() # 6D → 3D projection
# Mathematical derivations
derivation = await math.derive("Derive the Polyakov action for the bosonic string")Generates research summaries from paper collections.
from stringtheory.agents.synthesizer import Synthesizer
synth = Synthesizer(LLMConfig(), output_dir="output/reviews")
# Generate review
summary = await synth.synthesize(papers, "holographic entanglement")
# Identify trends
trends = await synth.identify_trends(papers)
# Compare approaches
comparison = await synth.compare_approaches(group_a, group_b, "quantum gravity")StringTheoryAI generates interactive 3D visualizations of key mathematical objects:
| Visualization | What It Shows |
|---|---|
| Calabi-Yau Manifold | 3D projection of the compact 6D spaces where extra dimensions live |
| String Worldsheet | The 2D surface swept by a vibrating string with oscillation modes |
| T² Torus | Toroidal compactification — the simplest extra-dimension geometry |
| Anti-de Sitter Space | The curved spacetime of AdS/CFT, showing the warp factor |
| D-Brane Stack | Multiple D-branes in transverse space with quantum fluctuations |
| Extra Dimensions | 6 compact dimensions projected into 3D spacetime |
Generate all at once:
stringtheory visualize all --format html --resolution 300Launch the full Streamlit interface:
pip install stringtheoryai[ui]
stringtheory uiFeatures:
- Paper search with expandable results and inline analysis
- Interactive concept explorer with concept cards
- Real-time 3D visualization generation
- Literature review generator with markdown download
StringTheoryAI/
├── stringtheory/
│ ├── agents/
│ │ ├── scholar.py # arXiv search & paper analysis
│ │ ├── theorist.py # Concept explanation & Q&A
│ │ ├── mathematician.py # Math computation & visualization
│ │ └── synthesizer.py # Literature review generation
│ ├── math/
│ │ └── manifolds.py # Calabi-Yau, worldsheet, AdS computations
│ ├── visualizations/
│ │ └── plotters.py # Plotly 3D visualization generators
│ ├── config.py # Pydantic configuration
│ ├── models.py # Data models (Paper, Concept, etc.)
│ ├── cli.py # Rich CLI interface
│ └── web.py # Streamlit web dashboard
├── examples/ # Usage examples
├── tests/ # Test suite
└── pyproject.toml
See the examples/ directory:
- search_papers.py — Search and analyze arXiv papers
- visualize_manifolds.py — Generate all math visualizations
- literature_review.py — Create an AI literature review
- Multi-agent architecture (Scholar, Theorist, Mathematician, Synthesizer)
- arXiv integration with hep-th, gr-qc, math-ph
- 6 interactive 3D visualizations
- Rich CLI + Streamlit Web UI
- Literature review generation
- PDF paper parsing (full-text analysis beyond abstracts)
- LaTeX equation rendering in terminal
- Research graph visualization (paper citation networks)
- Collaborative research notebooks
- Integration with Semantic Scholar API
- Fine-tuned models for string theory
- Automated conjecture generation
- String landscape exploration tools
Contributions welcome! Whether you're a physicist, developer, or both:
- Fork the repository
- Create your branch:
git checkout -b feature/amazing-feature - Run tests:
pytest - Submit a pull request
@software{stringtheoryai2024,
title={StringTheoryAI: AI-Powered Research Agent for String Theory},
year={2024},
url={https://github.com/gewenbo888/StringTheoryAI},
license={MIT}
}MIT License — see LICENSE.