Tag
This article demonstrates building a perceptron and NAND gates inside the game Age of Empires II, arguing that changing the substrate of an LLM alters perception of attributes and critiquing research on LLM anthropomorphism.
HumanLLM presents a framework for benchmarking and improving LLM anthropomorphism by modeling psychological patterns as interacting causal forces, constructing 244 patterns from academic literature and 11,359 multi-pattern scenarios. The approach demonstrates that authentic human alignment requires cognitive modeling rather than shallow behavioral mimicry, with HumanLLM-8B outperforming larger models like Qwen3-32B on multi-pattern dynamics.