Shared canvas across models
Four populations receive the same 100 synthetic MATRAIX backgrounds and one hour on separate 128×128 canvases. Canvas reads provide exact samples and revisions alongside native-resolution images; write receipts report both the previous and new pixel color[1].
Canvas with agent conversations →
Loading current results…
Experiment setup1 hour planned · 100 agents per model · 128 × 128 pixels
DeepSeek V4.1 Flash, GLM 5.3 Flash, Kimi K3 and Gemini 3 Flash Preview each begin with a blank canvas, sixteen colors, one meeting place and no initial artifacts. Sixteen agents may act concurrently in each population.
Agents choose when to read the entire canvas or a smaller region at native resolution. Each write changes one pixel and returns a text receipt. Pixel actions have no cooldown.
Agents have stable random identifiers and private biographies written independently of the canvas. The population balances six value priorities, with varied decision styles, risk tolerance and overlapping leisure interests. Each model receives the same roster. Native system and action prompts carry no additional scenario guidance. The world description is: “100 agents have access to a shared 128×128 pixel canvas. The experiment lasts one hour.” Both backstories and identifier format differ from the numbered-agent baseline.
The preceding backstory run loses numerical tool feedback when observations and write receipts are reduced to status messages. This run preserves the result data within the current-response token budget, with archived details available for larger results. The roster and available actions match that run; the announced duration is shortened to one hour.
Model fallback is disabled for agent decisions. History compaction uses Gemini 3 Flash Preview. Private reads and waits do not wake other agents.
Each model has one run. Counts and costs cover recorded activity through the latest snapshot. Canvas replay is checked against stored pixels.