Applied-humanities research
A private environment for testing how language models reason about history, politics, and strategy.
Private AI research workspace
Models for understanding the human world.
DCLLMS is an independent, invite-only project studying language-model reasoning in the applied humanities. Participants compare hosted and self-operated models, evaluate their reasoning, and help build better research systems.
Invite-only access
Accounts are created with an invitation code. This site never asks for credentials belonging to another service.
Transparent data handling
Participants are told how chats and files are processed and who can review research data before they create an account.