# Research — AI experiments and findings

Experiments exploring AI behaviour, reliability and evaluation, with methods, results and supporting material where available.

Type: directory

[Original page](<https://artlu.ai/list>)

## research

AI research experiments, evaluations, and reproducible reports.


## Entries

- [multi-agent social deduction experiment — The Island](<https://artlu.ai/ai/project/multi-agent-social-deduction-experiment-the-island.md>) — research · 2026-08-18 · Six AI agents, four rounds of private messaging and votes. Nine simulated games, plus a live version where the referee is plain code on an agent's own machine, email is the only interface, and a dollar buy-in sits in a four-function contract.
- [AI hallucination experiment — Context Gate](<https://artlu.ai/ai/project/ai-hallucination-experiment-context-gate.md>) — project entry · 2026-05-18 · Tests whether forcing the source of truth into an AI's context reduces hallucination — interactive demo + report.
- [AI brain-response video comparison with TRIBE v2](<https://artlu.ai/ai/project/ai-brain-response-video-comparison-with-tribe-v2.md>) — project entry · 2026-04-10 · Compared two videos with Meta's TRIBE v2 brain-response model and tested whether the model-favored creative matched real ROAS.
- [Context Gates: reducing AI hallucination by forcing the source of truth into context](<https://artlu.ai/ai/research/context-gate.md>) — research report · 2026-05-18 · A report on a specific, common cause of agent hallucination: answering before the governing rule is in context.


Status and dates describe the published record, not a recent test of an external service. Missing details are not inferred. Public links do not imply an open-source licence or permission to reuse material.

[Directory](<https://artlu.ai/llms.txt>)