我们精心收集整理社交人工智能(Social AI)领域的研究论文,并持续更新论文的中文摘要,从而支持快速了解该领域的代表性工作。仓库将不断更新,追踪社交 AI 前沿。欢迎 Follow 和 Star!⭐

🎯 范畴与结构

Social AI 是研究与构建具备社会智能的 AI 系统的领域。社会智能指智能体在社交情境中的三组能力:社交理解(感知情绪、意图、信念、关系、规范与多方动态)、社交推理(心智建模、归因与策略规划)、社交行动(说服、谈判、共情支持、建立信任)。

本仓库按四层结构组织:

• 任务层 · 社交任务
  • Persuasion & Negotiation — 说服、谈判等策略性对话的研究、方法与评测
  • Empathy & Emotional Support — 共情对话、情感支持与心理辅导
  • Conversational Recommendation — 对话式推荐系统
  • Cooperation & Collaboration — 合作、协作与多方共识构建
• 能力层 · 社交认知
  • Theory of Mind — 心智建模、意图与信念推理
  • Emotion Understanding — 情绪识别、归因与情绪推理
  • Social Norms & Morality — 社会规范、道德判断与价值对齐
  • Agent Memory — 智能体长期记忆与经验演化
• 方法层 · 支撑技术
  • Reinforcement Learning & Alignment — 强化学习训练与对齐方法
• 资源层 · 研究基础设施
  • User Simulation & Interactive Environments — 用户模拟、交互环境与仿真
  • Benchmark & Evaluation — 基准测试、数据集与评测

不收录:计算社会科学 / "AI for social science" 类工作。方法层的通用强化学习与对齐技术,以"支撑社交智能体的基础方法"身份收录;与社交无关、亦不服务于社交智能体的技术不收录。

参考论文:

1. Towards Social AI: A Survey on Understanding Social Interactions
2. Advancing Social Intelligence in AI Agents: Technical Challenges and Open Questions

📑 目录

• 🌐 Awesome Social AI
  • 🎯 范畴与结构
  • 📑 目录
  • 📚 论文列表
    • Persuasion & Negotiation
    • Empathy & Emotional Support
    • Conversational Recommendation
    • Cooperation & Collaboration
    • Theory of Mind
    • Emotion Understanding
    • Social Norms & Morality
    • Agent Memory
    • Reinforcement Learning & Alignment
    • User Simulation & Interactive Environments
    • Benchmark & Evaluation
  • 📁 仓库结构
  • ✍️ 如何贡献
  • 🧠 关于我们

📚 论文列表

🗣️ Persuasion & Negotiation · 41 篇

年份	会议/期刊	论文	链接	摘要	代码
2008	PERSUASIVE	A Systematic Framework for Designing and Evaluating Persuasive Systems	查看	摘要	-
2008	-	Measure Of Belief Change as an Evaluation of Persuasion	查看	摘要	-
2014	COLING	Reinforcement Learning of Cooperative Persuasive Dialogue Policies using Framing	查看	摘要	-
2017	HCI	Persuasive Argumentation and Emotions- An Empirical Evaluation with Users	查看	摘要	-
2023	ArgComp	Strategic argumentation dialogues for persuasion- Framework and experiments based on modelling the beliefs and concerns of the persuadee	查看	摘要	-
2023	arXiv	Improving Language Model Negotiation with Self-Play and In-Context Learning from AI Feedback	查看	摘要	-
2024	ICLR	Plug-and-Play Policy Planner for LLM-Powered Dialogue Agents	查看	摘要	代码
2024	ACL	How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs	查看	-	代码
2024	ACL	Measuring Bargaining Abilities of LLMs: A Benchmark and A Buyer-Enhancement Method	查看	-	代码
2024	EMNLP	NegotiationToM: A Benchmark for Stress-testing Machine Theory of Mind on Negotiation Surrounding	查看	-	代码
2024	ICML	Debating with More Persuasive LLMs Leads to More Truthful Answers	查看	-	代码
2024	NeurIPS	Cooperation, Competition, and Maliciousness: LLM-Stakeholders Interactive Negotiation	查看	-	代码
2025	Science	Durably reducing conspiracy beliefs through dialogues with AI	查看	摘要	-
2025	Science	The Levers of Political Persuasion	查看	摘要	代码
2025	ACL	Battling against Tough Resister- Strategy Planning with Adversarial Game for Non-collaborative Dialogues	查看	摘要	-
2025	TACL	Human Choice Prediction in Language-based Persuasion Games- Simulation-based Off-Policy Evaluation	查看	摘要	代码
2025	EMNLP	Enhancing LLM-Based Persuasion Simulations with Cultural and Speaker-Specific Information	查看	摘要	代码
2025	EMNLP	Enhancing Persuasive Dialogue Agents by Synthesizing Cross-Disciplinary Communication Strategies	查看	摘要	-
2025	NAACL	Teaching models to balance resisting and accepting persuasion	查看	-	代码
2025	arXiv	LLM Can be a Dangerous Persuader- Empirical Study of Persuasion Safety in Large Language Models	查看	摘要	代码
2025	EMNLP	PRINCIPLES- Synthetic Strategy Memory for Proactive Dialogue Agents	查看	摘要	代码
2025	NeurIPS	Persuade Me if You Can- A Framework for Evaluating Persuasion Effectiveness and Susceptibility Among Large Language Models	查看	摘要	代码
2025	AAAI	Simulation-free hierarchical latent policy planning for proactive dialogues	查看	摘要	-
2025	arXiv	ToMAP: Training Opponent-Aware LLM Persuaders with Theory of Mind	查看	摘要	代码
2025	ACL	EPO: Explicit Policy Optimization for Strategic Reasoning in LLMs via RL	查看	摘要	-
2025	arXiv	Disagreements in Reasoning: How a Model's Thinking Process Dictates Persuasion in Multi-Agent Systems	查看	摘要	-
2025	arXiv	EvoEmo: Evolved Emotional Policies for Adversarial LLM Agents in Multi-Turn Price Negotiation	查看	摘要	-
2025	arXiv	From Simulation to Strategy: Automating Personalized Interaction Planning for Conversational Agents	查看	摘要	-
2025	arXiv	Persuasion Should be Double-Blind: A Multi-Domain Dialogue Dataset With Faithfulness Based on Causal Theory of Mind	查看	摘要	-
2025	arXiv	Verbalized Bayesian Persuasion	查看	摘要	-
2025	EMNLP	Persuasion-Dynamics in LLMs: Investigating Robustness and Adaptability in Knowledge and Safety with DuET-PD	查看	摘要	-
2025	EMNLP	Profiling LLM Copyright Infringement Risks under Adversarial Persuasive Prompting	查看	摘要	代码
2025	CSUR	Persuasive Conversational Agents for Environmental Sustainability: A Survey	查看	摘要	-
2025	ACL	ASTRO: Automatic Strategy Optimization For Non-Cooperative Dialogues	查看	摘要	代码
2026	arXiv	One Model, All Roles- Multi-Turn, Multi-Agent Self-Play Reinforcement Learning for Conversational Social Intelligence	查看	-	-
2026	arXiv	Personality-Aware Reinforcement Learning for Persuasive Dialogue with LLM-Driven Simulation	查看	摘要	-
2026	CSUR	A Comprehensive Survey of Computational Persuasion	查看	摘要	代码
2026	ICLR	RebuttalAgent: Strategic Persuasion in Academic Rebuttal via Theory of Mind	查看	摘要	代码
2026	ICLR	Towards Strategic Persuasion with Language Models	查看	摘要	-
2026	arXiv	METRO: Towards Strategy Induction from Expert Dialogue Transcripts for Non-collaborative Dialogues	查看	摘要	代码
2026	ICLR	Strategic Planning and Rationalizing on Trees Make LLMs Better Debaters	查看	-	代码

❤️‍🩹 Empathy & Emotional Support · 19 篇

年份	会议/期刊	论文	链接	摘要	代码
2021	ACL	Towards Emotional Support Dialog Systems	查看	-	代码
2023	arXiv	CharacterChat- Learning towards Conversational AI with Personalized Social Support	查看	摘要	代码
2023	EMNLP	SoulChat- Improving LLMs’ Empathy, Listening, and Comfort Abilities through Fine-tuning with Multi-turn Empathy Conversations	查看	摘要	代码
2023	ACL	TransESC: Smoothing Emotional Support Conversation via Turn-Level State Transition	查看	-	代码
2024	ACL	EmoBench- Evaluating the Emotional Intelligence of Large Language Models	查看	摘要	代码
2024	ACL	Can Large Language Models be Good Emotional Supporter? Mitigating Preference Bias on Emotional Support Conversation	查看	-	代码
2024	ACL	ESCoT: Towards Interpretable Emotional Support Dialogue Systems	查看	-	代码
2025	arXiv	Echo-N1- Affective RL Frontier	查看	摘要	-
2025	ACL	PsyDT- Using LLMs to Construct the Digital Twin of Psychological Counselor with Personalized Counseling Style for Psychological Counseling	查看	摘要	代码
2025	ACL	PsyDial- A Large-scale Long-term Conversational Dataset for Mental Health Support	查看	摘要	代码
2025	arXiv	Reinforcement Learning with Verifiable Emotion Rewards for Empathetic Agents	查看	摘要	代码
2025	arXiv	SAGE- Steering and Refining Dialog Generation with State-Action Augmentation	查看	摘要	代码
2025	ACL	Beyond Verbal Cues: Emotional Contagion Graph Network for Causal Emotion Entailment	查看	-	代码
2025	EMNLP	Chain of Strategy Optimization Makes Large Language Models Better Emotional Supporter	查看	摘要	代码
2025	CHI	Customizing Emotional Support: How Do Individuals Construct and Interact With LLM-Powered Chatbots	查看	-	-
2025	NAACL	EmoDynamiX: Emotional Support Dialogue Strategy Prediction by Modelling MiXed Emotions and Discourse Dynamics	查看	-	代码
2026	arXiv	Affective Flow Language Model for Emotional Support Conversation	查看	摘要	代码
2026	arXiv	EMPA: Evaluating Persona-Aligned Empathy as a Process	查看	摘要	代码
2026	ACL	You Never Know a Person, You Only Know Their Defenses- Detecting Levels of Psychological Defense Mechanisms in Supportive Conversations	查看	-	-

🛍️ Conversational Recommendation · 24 篇

年份	会议/期刊	论文	链接	摘要	代码
2020	ACL	Towards Conversational Recommendation over Multi-Type Dialogs	查看	摘要	代码
2021	ACL	RevCore- Review-augmented Conversational Recommendation	查看	摘要	代码
2023	KDD	Improving conversational recommendation systems via counterfactual data simulation	查看	摘要	代码
2023	EMNLP	Rethinking the Evaluation for Conversational Recommendation in the Era of Large Language Models	查看	摘要	代码
2023	CIKM	Large Language Models as Zero-Shot Conversational Recommenders	查看	-	代码
2024	EMNLP	Beyond Persuasion: Towards Conversational Recommender System with Credible Explanations	查看	摘要	代码
2024	WWW	How Reliable is Your Simulator- Analysis on the Limitations of Current LLM-based User Simulators for Conversational Recommendation	查看	摘要	代码
2024	ACL	LLM-REDIAL: A Large-Scale Dataset for Conversational Recommender Systems Created from User Behaviors with LLMs	查看	摘要	代码
2024	arXiv	Reindex-Then-Adapt- Improving Large Language Models for Conversational Recommendation	查看	摘要	-
2024	ACL	Pearl: A Review-driven Persona-Knowledge Grounded Conversational Recommendation Dataset	查看	-	代码
2024	EMNLP	Mitigating Matthew Effect: Multi-Hypergraph Boosted Multi-Interest Self-Supervised Learning for Conversational Recommendation	查看	-	代码
2024	SIGIR	Broadening the View: Demonstration-augmented Prompt Learning for Conversational Recommendation	查看	-	-
2025	TKDE	A Causal-Based Attribute Selection Strategy for Conversational Recommender Systems	查看	摘要	-
2025	WWW	Bridging Conversational and Collaborative Signals for Conversational Recommendation	查看	摘要	-
2025	WWW	Collaborative Retrieval for Large Language Model-based Conversational Recommender Systems	查看	摘要	代码
2025	WWW	Towards Efficient Conversational Recommendations- Expected Value of Information Meets Bandit Learning	查看	摘要	-
2025	arXiv	A Framework for Generating Conversational Recommendation Datasets from Behavioral Interactions	查看	摘要	-
2025	EMNLP	LLM-based Conversational Recommendation Agents with Collaborative Verbalized Experience	查看	摘要	代码
2025	EMNLP	Towards Personalized Conversational Sales Agents	查看	摘要	-
2025	NAACL	Empowering Retrieval-based Conversational Recommendation with Contrasting User Preferences	查看	-	代码
2026	WWW	Not All Information Brings Benefits- Personalization-Driven Agent Debate for Conversational Recommendation	查看	摘要	-
2026	WWW	Optimizing Multi-Turn Interactive Recommendation Agents via Generative Intrinsic Motivation	查看	摘要	代码
2026	arXiv	User Simulator-Guided Multi-Turn Preference Optimization for Reasoning LLM-based Conversational Recommendation	查看	摘要	-
2026	ICLR	Rank-GRPO: Training LLM-based Conversational Recommender Systems with Reinforcement Learning	查看	-	代码

🤝 Cooperation & Collaboration · 11 篇

年份	会议/期刊	论文	链接	摘要	代码
2022	Science	Human-level play in the game of Diplomacy by combining language models with strategic reasoning	查看	-	-
2022	CHI	AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Prompts	查看	-	-
2024	ICLR	Building Cooperative Embodied Agents Modularly with Large Language Models	查看	-	-
2024	TACL	Decision-Oriented Dialogue for Human-AI Collaboration	查看	-	-
2024	ICML	Should we be going MAD- A Look at Multi-Agent Debate Strategies for LLMs	查看	-	代码
2024	ACL	Your Co-Workers Matter- Evaluating Collaborative Capabilities of Language Models in Blocks World	查看	-	-
2024	ACL	Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View	查看	-	代码
2024	ICLR	MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework	查看	-	代码
2024	NeurIPS	Cooperate or Collapse: Emergence of Sustainable Cooperation in a Society of LLM Agents	查看	-	代码
2024	Science	AI can help humans find common ground in democratic deliberation	查看	-	代码
2025	Nature Human Behaviour	Playing repeated games with large language models	查看	-	代码

💭 Theory of Mind · 19 篇

年份	会议/期刊	论文	链接	摘要	代码
2011	CogSci	Bayesian Theory of Mind- Modeling Joint Belief-Desire Attribution	查看	摘要	-
2019	COBS	Theory of Mind as Inverse Reinforcement Learning	查看	摘要	-
2024	ACL	Think Twice: Perspective-Taking Improves Large Language Models' Theory-of-Mind Capabilities	查看	-	代码
2025	ACL	Machine Theory of Mind Needs Machine Validation	查看	摘要	-
2025	ACL	Theory of Mind in Large Language Models- Assessment and Enhancement	查看	摘要	-
2025	arXiv	MINDGAMES: Do Large Language Models Have a Planning Theory of Mind?	查看	摘要	代码
2025	arXiv	Modeling the Mental World for Embodied AI- A Comprehensive Review	查看	摘要	-
2025	arXiv	ToM-agent: Large Language Models as Theory of Mind Aware Generative Agents with Counterfactual Reflection	查看	摘要	-
2025	arXiv	ToM-RL: Reinforcement Learning Unlocks Theory of Mind in Small LLMs	查看	摘要	代码
2025	ICML	Overcoming Multi-step Complexity in Multimodal Theory-of-Mind Reasoning- A Scalable Bayesian Planner	查看	摘要	-
2025	NeurIPS	AutoToM- Scaling Model-based Mental Inference via Automated Agent Modeling	查看	摘要	-
2025	NeurIPS	MetaMind- Modeling Human Social Thoughts with Metacognitive Multi-Agent Systems	查看	摘要	代码
2026	AAAI	Reality vs Counterfactual- Multi-World Contrastive Reinforcement Learning for Enhancing MLLM’s Theory of Mind in Egocentric Videos	查看	摘要	-
2026	arXiv	Infusing Theory of Mind into Socially Intelligent LLM Agents	查看	摘要	代码
2026	arXiv	MetaMind- General and Cognitive World Models in Multi-Agent Systems by Meta-Theory of Mind	查看	摘要	-
2026	arXiv	MindClaw- Closed-Loop Embodied Mental-State Reasoning for Precision Intervention	查看	摘要	-
2026	arXiv	UserHarness- Harnessing User Minds for Stronger Agent Theory-of-Mind	查看	摘要	-
2026	CVPR	Video-Only ToM- Enhancing Theory of Mind in Multimodal Large Language Models	查看	摘要	代码
2026	EACL	Let's Put Ourselves in Sally's Shoes: Shoes of Others Prefilling Improves Theory of Mind in LLMs	查看	摘要	-

🎭 Emotion Understanding · 10 篇

年份	会议/期刊	论文	链接	摘要	代码
2019	AAAI	DialogueRNN- An Attentive RNN for Emotion Detection in Conversations	查看	-	-
2019	ACL	MELD- A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations	查看	-	-
2021	AAAI	COSMIC- COmmonSense knowledge for eMotion Identification in Conversations	查看	-	-
2023	arXiv	InstructERC- Reforming Emotion Recognition in Conversation with Multi-task Retrieval-Augmented Large Language Models	查看	-	-
2023	AAAI	Knowledge-Bridged Causal Interaction Network for Causal Emotion Entailment	查看	-	代码
2024	KDD	EmoLLMs: A Series of Emotional Large Language Models and Annotation Tools for Comprehensive Affective Analysis	查看	-	代码
2024	NAACL	TelME: Teacher-leading Multimodal Fusion Network for Emotion Recognition in Conversation	查看	-	代码
2024	NeurIPS	Emotion-LLaMA: Multimodal Emotion Recognition and Reasoning with Instruction Tuning	查看	-	代码
2025	arXiv	Do LLMs Feel- Teaching Emotion Recognition with Prompts, Retrieval, and Curriculum Learning	查看	-	-
2025	ACL	CoE: A Clue of Emotion Framework for Emotion Recognition in Conversations	查看	-	-

⚖️ Social Norms & Morality · 12 篇

年份	会议/期刊	论文	链接	摘要	代码
2020	EMNLP	Social Chemistry 101- Learning to Reason about Social and Moral Norms	查看	-	-
2021	arXiv	Delphi- Towards Machine Ethics and Norms	查看	-	-
2021	ICLR	Aligning AI With Shared Human Values	查看	-	代码
2022	ACL	The Moral Integrity Corpus- A Benchmark for Ethical Dialogue Systems	查看	-	代码
2022	NeurIPS	When to Make Exceptions: Exploring Language Models as Accounts of Human Moral Judgment	查看	-	代码
2023	ICML	Do the Rewards Justify the Means- Measuring Trade-Offs Between Rewards and Ethical Behavior in the MACHIAVELLI Benchmark	查看	-	-
2023	ACL	NormBank- A Knowledge Bank of Situational Social Norms	查看	-	代码
2023	NeurIPS	Evaluating the Moral Beliefs Encoded in LLMs	查看	-	代码
2024	arXiv	CultureBank- An Online Community-Driven Knowledge Base Towards Culturally Aware Language Technologies	查看	-	代码
2024	AAAI	Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties	查看	-	代码
2025	NAACL	NormAd: A Framework for Measuring the Cultural Adaptability of Large Language Models	查看	-	代码
2025	Science Advances	Emergent social conventions and collective bias in LLM populations	查看	-	代码

💾 Agent Memory · 11 篇

年份	会议/期刊	论文	链接	摘要	代码
2023	UIST	Generative Agents- Interactive Simulacra of Human Behavior	查看	摘要	代码
2023	arXiv	MemoryBank- Enhancing Large Language Models with Long-Term Memory	查看	-	代码
2023	NeurIPS	Reflexion: Language Agents with Verbal Reinforcement Learning	查看	-	代码
2024	AAAI	ExpeL: LLM Agents Are Experiential Learners	查看	-	代码
2024	NeurIPS	HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models	查看	-	代码
2025	arXiv	A-Mem: Agentic Memory for LLM Agents	查看	摘要	代码
2025	arXiv	Evo-Memory- Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory	查看	摘要	代码
2025	arXiv	Remember Me, Refine Me- A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution	查看	摘要	代码
2025	ICLR	Human-inspired Episodic Memory for Infinite Context LLMs	查看	-	代码
2026	ICLR	MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent	查看	摘要	-
2026	ICLR	ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory	查看	摘要	代码

🤖 Reinforcement Learning & Alignment · 18 篇

年份	会议/期刊	论文	链接	摘要	代码
2022	NeurIPS	The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games	查看	-	代码
2022	NeurIPS	Training language models to follow instructions with human feedback	查看	-	-
2023	NeurIPS	Direct Preference Optimization- Your Language Model is Secretly a Reward Model	查看	摘要	-
2024	ICML	RLAIF vs. RLHF- Scaling Reinforcement Learning from Human Feedback with AI Feed	查看	摘要	-
2024	ICLR	Let's Verify Step by Step	查看	-	代码
2024	ICML	ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RL	查看	-	代码
2024	ICML	KTO: Model Alignment as Prospect Theoretic Optimization	查看	-	代码
2025	arXiv	DCPO- Dynamic Clipping Policy Optimization	查看	摘要	代码
2025	arXiv	Group Sequence Policy Optimization	查看	摘要	-
2025	COLING	MCA-Model-Based Causal RL for Efficient Dialogue Policy	查看	摘要	-
2025	NeurIPS	World Models Should Prioritize the Unification of Physical and Social Dynamics	查看	摘要	-
2025	EMNLP	Dream to Chat: Model-based Reinforcement Learning on Dialogues with User Belief Modeling	查看	摘要	-
2025	arXiv	Enhancing User Engagement in Socially-Driven Dialogue through Interactive LLM Alignments	查看	摘要	-
2025	arXiv	MAPO: Mixed Advantage Policy Optimization for Long-Horizon Multi-Turn Dialogue	查看	摘要	-
2025	ICML	VinePPO: Refining Credit Assignment in RL Training of LLMs	查看	-	代码
2026	ICLR	Toward Evaluative Thinking: Meta-Policy Optimization with Evolving Reward Models	查看	摘要	-
2026	arXiv	Better LLM Reasoning via Dual-Play	查看	摘要	代码
2026	ICLR	TreeSearch for LLM Agent Reinforcement Learning	查看	摘要	代码

🕹️ User Simulation & Interactive Environments · 16 篇

年份	会议/期刊	论文	链接	摘要	代码
2006	KER	A Survey of Statistical User Simulation Techniques for RL Dialogue Management	查看	摘要	-
2023	TOIS	Metaphorical User Simulators for Evaluating Task-oriented Dialogue Systems	查看	摘要	代码; 代码
2024	ICLR	SOTOPIA- Interactive Evaluation for Social Intelligence in Language Agents	查看	摘要	-
2024	WWW	An In-depth Investigation of User Response Simulation for Conversational Search	查看	摘要	代码
2024	AAAI	Adversarial Socialbots Modeling Based on Structural Information Principles	查看	摘要	代码
2024	arXiv	Strength Lies in Differences! Improving Strategy Planning for Non-collaborative Dialogues via Diversified User Simulation	查看	摘要	-
2024	ACL	PlatoLM: Teaching LLMs in Multi-Round Dialogue via a User Simulator	查看	-	代码
2024	arXiv	LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals	查看	-	代码
2024	arXiv	OASIS: Open Agent Social Interaction Simulations with One Million Agents	查看	-	代码
2025	SIGDIAL	Generating Diverse Personas for User Simulators to Test Interview Dialogue Systems	查看	摘要	-
2025	SIGIR	Simulating Before Planning- Constructing Intrinsic User World Model for User-Tailored Dialogue Policy Planning	查看	摘要	-
2025	SIGIR	Theory and Toolkits for User Simulation in the Era of Generative AI- User Modeling, Synthetic Data Generation, and System Evaluation	查看	摘要	-
2025	WWW	A LLM-based Controllable, Scalable, Human-Involved User Simulator Framework for Conversational Recommender Systems	查看	摘要	代码
2025	NeurIPS	Goal Alignment in LLM-Based User Simulators for Conversational AI	查看	摘要	代码
2025	ICLR	τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains	查看	-	代码
2026	CSUR	From Individual to Society: A Survey on Social Simulation Driven by Large Language Model-based Agents	查看	-	代码

📊 Benchmark & Evaluation · 33 篇

年份	会议/期刊	论文	链接	摘要	代码
2019	CVPR	Social-IQ- A Question Answering Benchmark for Artificial Social Intelligence	查看	-	-
2023	ICCV	Social-IQ 2.0 Challenge- Benchmarking Multimodal Social Understanding	查看	-	代码
2023	EMNLP	FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions	查看	-	代码
2023	NeurIPS	Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena	查看	-	代码
2023	NeurIPS	Understanding Social Reasoning in Language Models with Language Models	查看	-	代码
2024	ACL	Evaluating Intention Detection Capability of Large Language Models in Persuasive Dialogues	查看	摘要	代码
2024	ICML	Agent-as-a-Judge- Evaluate Agents with Agents	查看	摘要	代码
2024	ACL	Evaluating Very Long-Term Conversational Memory of LLM Agents	查看	-	代码
2024	ACL	MM-SOC: Benchmarking Multimodal Large Language Models in Social Media Platforms	查看	-	代码
2024	ACL	MMToM-QA: Multimodal Theory of Mind Question Answering	查看	-	代码
2024	ACL	OpenToM: A Comprehensive Benchmark for Evaluating Theory-of-Mind Reasoning Capabilities of Large Language Models	查看	-	代码
2024	ACL	SocialBench: Sociality Evaluation of Role-Playing Conversational Agents	查看	-	代码
2024	ACL	ToMBench: Benchmarking Theory of Mind in Large Language Models	查看	-	代码
2024	ICLR	Who is ChatGPT? Benchmarking LLMs' Psychological Portrayal Using PsychoBench	查看	-	代码
2025	AAAI	MuMA-ToM- Multi-modal Multi-Agent Theory of Mind	查看	摘要	代码
2025	AAAI	ToMATO: Verbalizing the Mental States of Role-Playing LLMs for Benchmarking Theory of Mind	查看	摘要	代码
2025	ACL	Towards Dynamic Theory of Mind- Evaluating LLM Adaptation to Temporal Evolution of Human States	查看	摘要	代码
2025	EMNLP	MOMENT S- A Comprehensive Multimodal Benchmark for Theory of Mind	查看	摘要	代码
2025	ICLR	Explore theory of mind: program-guided adversarial data generation for theory of mind reasoning	查看	摘要	代码
2025	NAACL	Communication Makes Perfect: Persuasion Dataset Construction via Multi-LLM Communication	查看	摘要	代码
2025	ACL	DICE-BENCH- Evaluating the Tool-Use Capabilities of Large Language Models in Multi-Round, Multi-Party Dialogues	查看	-	代码
2025	ACL	In Search of the Lost Arch in Dialogue- A Dependency Dialogue Acts Corpus for Multi-Party Dialogues	查看	-	-
2025	NAACL	WHoW- A Cross-domain Approach for Analysing Conversation Moderation	查看	-	-
2025	arXiv	You need to MIMIC to get FAME- Solving Meeting Transcript Scarcity with Multi-Agent Conversations	查看	-	-
2025	EMNLP	PersonaGym: Evaluating Persona Agents and LLMs	查看	-	代码
2025	ICLR	LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory	查看	-	代码
2026	AAAI	RecToM: A Benchmark for Evaluating Machine Theory of Mind in LLM-based Conversational Recommender Systems	查看	摘要	代码
2026	arXiv	EnactToM- An Evolving Benchmark for Functional Theory of Mind in Embodied Agents	查看	摘要	代码
2026	arXiv	Large language model psychometrics- A systematic review of evaluation, validation, and enhancement	查看	摘要	代码
2026	WWW	ES-MemEval- Benchmarking Conversational Agents on Personalized Long-Term Emotional Support	查看	摘要	代码
2026	arXiv	PIVOTSBench- Evaluating Fine-Grained Interpersonal Relationship Reasoning in Multimodal Large Language Models	查看	-	-
2026	arXiv	TIDES- A Longitudinal Bilingual Dataset for Modeling Multi-Party Social Dynamics	查看	-	-
2026	ICLR	MME-Emotion: A Holistic Evaluation Benchmark for Emotional Intelligence in Multimodal Large Language Models	查看	-	代码

---

📁 仓库结构

Awesome-Social-AI/
├── image/              # 存放论文相关的图片、图表等
├── paper/              # 论文摘要（main，按 8 个方向前缀命名）
├── Example.md          # 论文摘要的撰写范例
└── README.md           # 本说明文件

---

✍️ 如何贡献

我们鼓励所有成员积极贡献自己阅读的论文摘要，请严格遵循以下步骤：

1. Fork & Clone

Fork 本仓库到你的 GitHub 账户，然后克隆到本地。

git clone https://github.com/你的用户名/Awesome-Social-AI.git && cd Awesome-Social-AI

2. 添加原始仓库为 upstream

git remote add upstream https://github.com/lucianma05-create/Awesome-Social-AI.git

3. 每次准备撰写新摘要前，请先同步主分支并创建一个独立分支：

1. 切换回主分支并拉取上游最新代码

git checkout main
git pull upstream main

2. 创建并切换到一个新分支 (分支名建议反映论文内容)

git checkout -b 分支名

4. 命名规范

\paper 下的文件名请严格按照以下格式命名，以便于检索和管理：
[方向]-[会议/期刊名]-[年份]-[论文名(完整的名字而不是缩写)].md
示例: Memory-NeurIPS-2023-Retrieval-Augmented-Generation.md
研究方向请从当前 8 个方向中选择最接近的归类；确需新增方向时，请先在组内讨论后再添加。
当前方向前缀：PD、ED、Recommend、Coop、ToM、Emotion、Norms、Memory、RLHF、US、Data。
\image 下的文件命名为 [年]-[月]-[日]-[编号]-[姓名缩写].png
示例：2024010101mmh.png

5. 填写内容

a. 参照 Example.md 中的模板，填写论文的各项信息，确保内容精炼、准确。

b. 或者可以使用我们专门开发的 Auto-Summary, 请在自动化生成后进行必要的人工校对和修改。

6. 修改 README.md

参照 README.md 中的表格，增加新论文的链接、摘要和代码；可以用以下提示词提示codex或copilot进行自动化填充：

请根据 paper 目录新增的 .md 文件，按 README.md 里现有表格格式补全相应方向的行，填 链接、摘要、代码 字段，缺失用 - 

或运行仓库自带的同步脚本，自动把 paper/ 目录的新论文填入表格：

python scripts/update_readme.py --dry-run   # 预览改动
python scripts/update_readme.py             # 应用改动

7. 提交、同步与推送

1. 暂存并提交本地更改

git add .
git commit -m "你的提交信息，例如：Add summary for [论文名]"

2. 推送当前功能分支到你个人的 GitHub (origin)

git push origin 分支名

8. 发起 Pull Request

向本仓库的主分支发起一个 Pull Request (PR)，并等待审核合并。

---

🧠 关于我们

本仓库由 NWPU Crowd-HMT-Lab Social-AI-Group 维护。