Caesar is an autonomous AI research agent. Instead of summarizing a flat list of search results, it treats the web as a graph — building a dynamic knowledge graph as it explores, backtracking when it stagnates, and refining its answer through an adversarial Generator–Verifier loop. The result is deeper, more novel synthesis on the open-ended, cross-disciplinary questions retrieval alone cannot answer.
Live site: https://jasonzliang.github.io/caesar-agent/
This repository hosts the public landing page only. It is built and served via GitHub Pages.
Today's deep-research agents — ChatGPT Deep Research, Perplexity, Gemini Deep Research, GPT Researcher — optimize retrieval precision over a flat sequence of documents. They produce competent summaries but fall into local minima, suffer from navigational amnesia, and converge on derivative, consensus-driven outputs.
Caesar is built differently:
- Builds a knowledge graph as it explores — each new page is analyzed against insights already attached to predecessor and neighbor nodes.
- Adversarial self-critique on its own draft — an independent verifier formulates orthogonal queries that target weaknesses in the current draft, escaping the consensus basin that traps single-pass LLMs.
- Multiple drafts, then merged into one — each draft chains off the previous one until a final merge.
- Backtracks when an exploration path stalls — depth-first drill-down with a stack to pop back and explore orthogonal branches.
- Multi-provider — OpenAI, Anthropic, Google Gemini, or any OpenAI-compatible endpoint.
- Reproducible run logs (JSON) — tokens, cost, wall-time, pages visited, per-draft provenance.
Blinded 3-model LLM-as-a-Judge panel (Claude Sonnet 4.5, GPT-5.2, Gemini 3 Pro) scored 0–10 across three creativity dimensions: New, Useful, Surprising.
| Agent | New | Useful | Surprising | Total |
|---|---|---|---|---|
| Caesar | 9.11 | 8.87 | 8.98 | 26.96 |
| Gemini 3 Deep Research | 8.09 | 7.60 | 8.09 | 23.78 |
| Sonnet 4.5 Deep Research | 6.73 | 7.49 | 6.42 | 20.64 |
| GPT-5.2 Deep Research | 5.07 | 6.31 | 4.36 | 15.74 |
Cliff's Delta effect sizes are uniformly large (δ ≥ 0.76, well above the 0.47 large-effect threshold) across all baselines and output formats; δ = 1.00 against five of six baselines indicates strict dominance. The advantage holds in a compute-controlled run (Caesar at $5/challenge with GPT-5-mini still tops Gemini 3 Deep Research) and in a 23-rater human A/B study (Caesar preferred 56.25%, odds ratio 1.29). Ablations confirm both graph exploration and the adversarial verifier loop are independently necessary. See the paper for full methodology, exploration-budget ablation, and judge bias analysis.
- Landing page (full details, figures, FAQ): https://jasonzliang.github.io/caesar-agent/
- Paper (arXiv): https://arxiv.org/abs/2604.20855
- DOI: https://doi.org/10.48550/arXiv.2604.20855
- ResearchGate: Caesar publication
@misc{liang26caesar,
title={Caesar: Deep Agentic Web Exploration for Creative Answer Synthesis},
author={Jason Liang and Elliot Meyerson and Risto Miikkulainen},
year={2026},
eprint={2604.20855},
archivePrefix={arXiv},
primaryClass={cs.IR},
url={https://arxiv.org/abs/2604.20855}
}By Jason Liang, Elliot Meyerson, and Risto Miikkulainen — Cognizant AI Lab and The University of Texas at Austin.
Apache License 2.0.