Paper Detail

Socio-Foundation: A Model for Generalizable Individual Behavior Simulation via Hierarchical Capability Distillation

Liang Wang, Wenxuan Xie, Xinyi Mou, Yixin Luo, Zhongyu Wei

arxiv Score 6.8

Published 2026-10-06 · First seen 2026-10-08

General AI

Abstract

Simulating individual behavior requires large language models (LLMs) to preserve persona traits while adapting to dynamic social contexts. However, general-purpose LLMs often flatten distinct personas, while task-specific tuning suffers from fragmentation and generalization. To overcome these challenges, we organize individual simulation into the \textbf{FONTS Taxonomy}, comprising five complementary capability dimensions: \emph{persona fidelity} (\textbf{F}), \emph{outcome realization} (\textbf{O}), \emph{behavioral naturalness} (\textbf{N}), \emph{trajectory coherence} (\textbf{T}), and \emph{social grounding} (\textbf{S}). Grounded in this taxonomy, we curate a standardized training corpus library of approximately 10 million instances across 14 representative datasets and present \textbf{Socio-Foundation}. Socio-Foundation decouples specialization from integration via a three-stage pipeline: learning task experts via DAPO, consolidating them into capability experts via off-policy distillation, and unifying them via multi-teacher on-policy distillation (MOPD). We also establish \textbf{IndiEval}, consolidating 29 metrics across the FONTS dimensions. Experiments show that Socio-Foundation outperforms its \textit{Qwen3-8B} base by 11.0 points and approaches frontier models such as \textit{GLM-5.2}, with ablations and out-of-distribution evaluations further demonstrating the effectiveness and generalization of our model.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{wang2026socio,
  title = {Socio-Foundation: A Model for Generalizable Individual Behavior Simulation via Hierarchical Capability Distillation},
  author = {Liang Wang and Wenxuan Xie and Xinyi Mou and Yixin Luo and Zhongyu Wei},
  year = {2026},
  abstract = {Simulating individual behavior requires large language models (LLMs) to preserve persona traits while adapting to dynamic social contexts. However, general-purpose LLMs often flatten distinct personas, while task-specific tuning suffers from fragmentation and generalization. To overcome these challenges, we organize individual simulation into the \textbackslash{}textbf\{FONTS Taxonomy\}, comprising five complementary capability dimensions: \textbackslash{}emph\{persona fidelity\} (\textbackslash{}textbf\{F\}), \textbackslash{}emph\{outcome realization\} (\textbackslash{}textbf},
  url = {https://arxiv.org/abs/2610.08967},
  keywords = {cs.AI},
  eprint = {2610.08967},
  archiveprefix = {arXiv},
}

Metadata

{}