SOTOPIA-$Ω$: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Wenyuan, Liu, Tianyun, Song, Mengxiao, Li, Xiaodong, Liu, Tingwen
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909626803421184
author Zhang, Wenyuan
Liu, Tianyun
Song, Mengxiao
Li, Xiaodong
Liu, Tingwen
author_facet Zhang, Wenyuan
Liu, Tianyun
Song, Mengxiao
Li, Xiaodong
Liu, Tingwen
contents Despite the abundance of prior social strategies possessed by humans, there remains a paucity of research dedicated to their transfer and integration into social agents. Our proposed SOTOPIA-$Ω$ framework aims to address and bridge this gap, with a particular focus on enhancing the social capabilities of language agents. This framework dynamically injects multi-step reasoning strategies inspired by negotiation theory and two simple direct strategies into expert agents, thereby automating the construction of a high-quality social dialogue training corpus. Additionally, we introduce the concept of Social Instruction Following (S-IF) and propose two new S-IF evaluation metrics that complement social capability. We demonstrate that several 7B models trained on high-quality corpus not only significantly surpass the expert agent (GPT-4) in achieving social goals but also enhance S-IF performance. Analysis and variant experiments validate the advantages of dynamic construction, which can especially break the agent's prolonged deadlock.
format Preprint
id arxiv_https___arxiv_org_abs_2502_15538
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SOTOPIA-$Ω$: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents
Zhang, Wenyuan
Liu, Tianyun
Song, Mengxiao
Li, Xiaodong
Liu, Tingwen
Computation and Language
Computers and Society
Human-Computer Interaction
Despite the abundance of prior social strategies possessed by humans, there remains a paucity of research dedicated to their transfer and integration into social agents. Our proposed SOTOPIA-$Ω$ framework aims to address and bridge this gap, with a particular focus on enhancing the social capabilities of language agents. This framework dynamically injects multi-step reasoning strategies inspired by negotiation theory and two simple direct strategies into expert agents, thereby automating the construction of a high-quality social dialogue training corpus. Additionally, we introduce the concept of Social Instruction Following (S-IF) and propose two new S-IF evaluation metrics that complement social capability. We demonstrate that several 7B models trained on high-quality corpus not only significantly surpass the expert agent (GPT-4) in achieving social goals but also enhance S-IF performance. Analysis and variant experiments validate the advantages of dynamic construction, which can especially break the agent's prolonged deadlock.
title SOTOPIA-$Ω$: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents
topic Computation and Language
Computers and Society
Human-Computer Interaction
url https://arxiv.org/abs/2502.15538