dorsal/arxiv
View SchemaHUMANLLM: Benchmarking and Reinforcing LLM Anthropomorphism via Human Cognitive Patterns
| Authors | Xintao Wang, Jian Yang, Weiyuan Li, Rui Xie, Jen-tse Huang, Jun Gao, Shuai Huang, Yueping Kang, Liyuan Gou, Hongwei Feng, Yanghua Xiao |
|---|---|
| Categories | |
| ArXiv ID | 2601.10198vv1 |
| URL | https://arxiv.org/abs/2601.10198 |
| License | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
Abstract
Large Language Models (LLMs) have demonstrated remarkable capabilities in reasoning and generation, serving as the foundation for advanced persona simulation and Role-Playing Language Agents (RPLAs). However, achieving authentic alignment with human cognitive and behavioral patterns remains a critical challenge for these agents. We present HUMANLLM, a framework treating psychological patterns as interacting causal forces. We construct 244 patterns from ~12,000 academic papers and synthesize 11,359 scenarios where 2-5 patterns reinforce, conflict, or modulate each other, with multi-turn conversations expressing inner thoughts, actions, and dialogue. Our dual-level checklists evaluate both individual pattern fidelity and emergent multi-pattern dynamics, achieving strong human alignment (r=0.91) while revealing that holistic metrics conflate simulation accuracy with social desirability. HUMANLLM-8B outperforms Qwen3-32B on multi-pattern dynamics despite 4x fewer parameters, demonstrating that authentic anthropomorphism requires cognitive modeling--simulating not just what humans do, but the psychological processes generating those behaviors.
{
"annotation_id": "7136d785-9c8f-4230-93f8-c96945118f5a",
"date_created": "2026-02-17T05:53:24.273000Z",
"date_modified": "2026-02-17T05:53:24.273000Z",
"file_hash": "c833439ebc161e2698ee1b8c1a9a15bffeb9d3751ff8d5bc5503119ad9165b0e",
"private": false,
"record": {
"abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities in reasoning and generation, serving as the foundation for advanced persona simulation and Role-Playing Language Agents (RPLAs). However, achieving authentic alignment with human cognitive and behavioral patterns remains a critical challenge for these agents. We present HUMANLLM, a framework treating psychological patterns as interacting causal forces. We construct 244 patterns from ~12,000 academic papers and synthesize 11,359 scenarios where 2-5 patterns reinforce, conflict, or modulate each other, with multi-turn conversations expressing inner thoughts, actions, and dialogue. Our dual-level checklists evaluate both individual pattern fidelity and emergent multi-pattern dynamics, achieving strong human alignment (r=0.91) while revealing that holistic metrics conflate simulation accuracy with social desirability. HUMANLLM-8B outperforms Qwen3-32B on multi-pattern dynamics despite 4x fewer parameters, demonstrating that authentic anthropomorphism requires cognitive modeling--simulating not just what humans do, but the psychological processes generating those behaviors.",
"arxiv_id": "2601.10198",
"authors": [
"Xintao Wang",
"Jian Yang",
"Weiyuan Li",
"Rui Xie",
"Jen-tse Huang",
"Jun Gao",
"Shuai Huang",
"Yueping Kang",
"Liyuan Gou",
"Hongwei Feng",
"Yanghua Xiao"
],
"categories": [
"cs.CL"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"title": "HUMANLLM: Benchmarking and Reinforcing LLM Anthropomorphism via Human Cognitive Patterns",
"url": "https://arxiv.org/abs/2601.10198",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "255e700c-247e-405c-bb0e-8f58742bbbd3",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}