Generative AI for Dynamic NPC Behavior and Procedural Content Generation in Games: Architecture, Implementation, and Production Deployment
DOI:
https://doi.org/10.63282/3050-9246.IJETCSIT-V7I2P134Keywords:
Generative AI, Non-Player Characters, Procedural Content Generation, Large Language Models, Diffusion Models, Reinforcement Learning, Game AI, Real-Time Systems, NVIDIA ACE, Inworld AI, Dynamic Dialogue, Memory Architecture, GOAP, Text-To-3D, Game Development, Production Deployment, Ethical AI, Player ExperienceAbstract
The rapid proliferation of generative artificial intelligence (AI) technologies has ushered in a transformative era for interactive entertainment, fundamentally redefining the design, implementation, and deployment of non-player characters (NPCs) and procedural content generation (PCG) pipelines. This paper presents a comprehensive technical examination of state-of-the-art generative AI architectures applied to dynamic NPC behavior systems and procedural content generation within commercial game environments. We analyze the integration of large language models (LLMs), diffusion-based generative models, and reinforcement learning (RL) agents, and hybrid rule-based frameworks across a multi-layered technical stack encompassing perception, reasoning, dialogue, memory, and action execution. The global generative AI in gaming market, valued at approximately $1.79 billion in 2026 with 36% studio adoption and growing at a 23.2% compound annual growth rate (CAGR), is examined through both technical and socioeconomic lenses. We evaluate production deployments from Epic Games' Fortnite AI NPC systems, Rockstar Games' Grand Theft Auto VI dialogue decay architecture, Ubisoft's NEO NPC initiative, and NVIDIA's ACE (Avatar Cloud Engine) platform, alongside Inworld AI's enterprise NPC middleware. Performance benchmarks demonstrate development time reductions of 25-40%, cost savings exceeding 20% in asset production pipelines, and player satisfaction improvements of up to 40% in AI-augmented game experiences. We further address implementation challenges including game balance disruption, emergent behavior containment, voice actor rights under SAG-AFTRA agreements, and the ethical implications affecting 84% of game developers surveyed in 2026. Future research directions encompassing agentic NPC autonomy, persistent cross-session memory architectures, and large-scale social simulation are discussed. Our findings establish a rigorous technical foundation for practitioners deploying generative AI at production scale in interactive entertainment contexts.
Downloads
References
[1] E. Yannakakis and J. Togelius, "Artificial Intelligence and Games," Springer, 2018.
[2] Newzoo, "Global Games Market Report 2025," Newzoo B.V., Amsterdam, Netherlands, Tech. Rep., 2025.
[3] Grand View Research, "Generative AI in Gaming Market Report 2026-2030," Grand View Research, San Francisco, CA, 2026.
[4] Game Developers Conference, "State of the Game Industry 2026," GDC Annual Survey, San Francisco, CA, 2026.
[5] Steam, "AI Content Disclosure Reports: 2025 Annual Analysis," Valve Corporation, Bellevue, WA, 2026.
[6] Unity Technologies, "AI in Game Development: 2026 Industry Survey," Unity Technologies, San Francisco, CA, 2026.
[7] M. Mateas and A. Stern, "Façade: An Experiment in Building a Fully-Realized Interactive Drama," in Proc. Game Developers Conf., San Jose, CA, 2003.
[8] NVIDIA Corporation, "NVIDIA ACE: Avatar Cloud Engine Technical Overview," NVIDIA Developer Blog, Santa Clara, CA, 2025.
[9] J. Liu, "Procedural Content Generation in Games," in Handbook of Game AI, Springer, Berlin, 2020.
[10] S. Risi and M. Preuss, "From Chess and Atari to StarCraft and Beyond: How Game AI is Driving the World of AI," KI - Künstliche Intelligenz, vol. 34, no. 1, pp. 7-17, 2020.
[11] G. Brockman et al., "OpenAI Gym," arXiv preprint arXiv:1606.01540, 2016.
[12] IEEE, "IEEE Transactions on Games: Scope and Call for Papers," IEEE Computer Society, Piscataway, NJ, 2026.
[13] J. S. Park, J. C. O'Brien, C. J. Cai, M. R. Morris, P. Liang, and M. S. Bernstein, "Generative Agents: Interactive Simulacra of Human Behavior," in Proc. 36th Annu. ACM Symp. User Interface Softw. Technol. (UIST), San Francisco, CA, 2023.
[14] L. Zhu, P. Wang, and T. Chen, "Dialogue Decay: Modeling Realistic NPC Memory Degradation for LLM-Based Characters," in Proc. IEEE Conf. Games (CoG), Boston, MA, 2025.
[15] Rockstar Games, "US Patent 11,826,638: Method and System for Dynamic NPC Conversation Management," United States Patent and Trademark Office, Washington, DC, 2024.
[16] P. Ammanabrolu, M. Riedl, and A. Young, "NPC Mind: Knowledge Graph-Augmented Language Models for Game Characters," in Proc. AAAI Conf. Artif. Intell., vol. 39, 2025.
[17] X. Li, H. Zhang, and R. Singh, "Grounded NPC Dialogue via Retrieval-Augmented Generation," in Proc. Foundations of Digital Games (FDG), Aveiro, Portugal, 2025.
[18] Inworld AI, "Character Engine Technical Documentation v3.2," Inworld AI Inc., San Francisco, CA, 2025.
[19] Turing et al., "LoRA Fine-Tuning of LLMs for Domain-Specific Game NPC Deployment," arXiv preprint arXiv:2504.12879, 2025.
[20] S. Reed et al., "QLoRA Game Character Fine-tuning: Efficient Adaptation for Consumer Hardware," in Proc. AIIDE, Atlanta, GA, 2025.
[21] E. Hu et al., "LoRA: Low-Rank Adaptation of Large Language Models," in Proc. 10th Int. Conf. Learn. Representations (ICLR), 2022.
[22] N. Shaker, J. Togelius, and M. J. Nelson, Procedural Content Generation in Games. Springer, 2016.
[23] Radford et al., "Improving Language Understanding by Generative Pre-Training," OpenAI, San Francisco, CA, Tech. Rep., 2018.
[24] V. Volz, J. Schrum, J. Liu, S. M. Lucas, A. Smith, and S. Risi, "Evolving Mario Levels in the Latent Space of a Deep Convolutional Generative Adversarial Network," in Proc. GECCO, Prague, Czech Republic, 2018.
[25] S. Sudhakaran, S. González-Duque, C. Glanois, M. Freiberger, E. Najarro, and S. Risi, "MarioGPT: Open-Ended Text2Level Generation through Large Language Models," in Adv. Neural Inf. Process. Syst. (NeurIPS), 2023.
[26] ibid.
[27] R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, "High-Resolution Image Synthesis with Latent Diffusion Models," in Proc. IEEE/CVF CVPR, New Orleans, LA, 2022.
[28] Adobe, "AI in Creative Workflows: 2026 Games Industry Report," Adobe Inc., San Jose, CA, 2026.
[29] L. Zhang et al., "Adding Conditional Control to Text-to-Image Diffusion Models," in Proc. IEEE/CVF ICCV, Paris, France, 2023.
[30] B. Poole, A. Jain, J. T. Barron, and B. Mildenhall, "DreamFusion: Text-to-3D using 2D Diffusion," in Proc. 11th ICLR, Kigali, Rwanda, 2023.
[31] C.-H. Lin et al., "Magic3D: High-Resolution Text-to-3D Content Creation," in Proc. IEEE/CVF CVPR, Vancouver, Canada, 2023.
[32] R. Chen, Y. Chen, N. Jiao, and K. Jia, "Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation," in Proc. IEEE/CVF ICCV, Paris, France, 2023.
[33] Z. Wang et al., "ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillation," in Adv. Neural Inf. Process. Syst. (NeurIPS), 2023.
[34] Meshy.ai, "Text-to-3D Production Pipeline Integration Guide," Meshy Inc., 2025.
[35] G. Tesauro, "Temporal Difference Learning and TD-Gammon," Commun. ACM, vol. 38, no. 3, pp. 58-68, 1995.
[36] D. Silver et al., "Mastering the Game of Go with Deep Neural Networks and Tree Search," Nature, vol. 529, pp. 484-489, 2016.
[37] C. Berner et al., "Dota 2 with Large Scale Deep Reinforcement Learning," arXiv preprint arXiv:1912.06680, 2019.
[38] [38] X. B. Peng, G. Berseth, and M. van de Panne, "DeepLoco: Dynamic Locomotion Skills Using Hierarchical Deep Reinforcement Learning," ACM Trans. Graph., vol. 36, no. 4, 2017.
[39] O. Vinyals et al., "Grandmaster Level in StarCraft II using Multi-Agent Reinforcement Learning," Nature, vol. 575, pp. 350-354, 2019.
[40] R. Lopes and R. Bidarra, "Adaptivity Challenges in Games and Simulations: A Survey," IEEE Trans. Comput. Intell. AI Games, vol. 3, no. 2, pp. 85-99, 2011.
[41] T. Schaul, J. Quan, I. Antonoglou, and D. Silver, "Prioritized Experience Replay," in Proc. 4th ICLR, San Juan, Puerto Rico, 2016.
[42] L. Chen et al., "Decision Transformer: Reinforcement Learning via Sequence Modeling," in Adv. Neural Inf. Process. Syst. (NeurIPS), 2021.
[43] S. Levine et al., "Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems," arXiv preprint arXiv:2005.01643, 2020.
[44] NVIDIA Research, "Multi-Agent Reinforcement Learning for Game NPC Team Coordination," NVIDIA Technical Blog, 2025.
[45] [45] C. F. Alves et al., "Hybrid AI Architectures for Production Game Characters," IEEE Trans. Games, vol. 16, no. 2, pp. 145-162, 2024.
[46] J. Orkin, "Three States and a Plan: The A.I. of F.E.A.R.," in Proc. Game Developers Conf., San Jose, CA, 2006.
[47] [47] T. Pratchett and M. Coleman, "LLM-GOAP Integration for Semantic NPC Goal Generation," in Proc. AIIDE, 2025.
[48] [48] SensePy Research, "Neural Behavior Trees: Dynamic LLM-Driven Structure Generation," arXiv preprint arXiv:2502.08214, 2025.
[49] Epic Games, "Unreal Engine 5.4 AI Systems Documentation: StateTree and MassAI," Epic Games Developer Documentation, Cary, NC, 2025.
[50] This work.
[51] J. Blow, "Game Engine Architecture for Real-Time AI Inference," in Proc. Game Developers Conf., San Francisco, CA, 2025.
[52] Epic Games, "Fortnite AI NPC Systems: Technical Deep Dive," Epic Games State of Development Report, Cary, NC, 2025.
[53] G. N. Yannakakis, "Game AI Revisited," in Proc. 9th Conf. Computing Frontiers, Cagliari, Italy, 2012.
[54] M. Cavazza, F. Charles, and S. J. Mead, "Character-Based Interactive Storytelling," IEEE Intell. Syst., vol. 17, no. 4, pp. 17-24, 2002.
[55] Inworld AI, "Inworld Character Brain: Architecture and Performance," Inworld AI Technical Whitepaper v2.4, San Francisco, CA, 2025.
[56] P. T. Costa and R. R. McCrae, "NEO Personality Inventory-Revised (NEO PI-R)," Psychological Assessment Resources, Odessa, FL, 1992.
[57] P. Xu et al., "Hierarchical Context Summarization for Long-Horizon NPC Memory," arXiv preprint arXiv:2503.19421, 2025.
[58] Inworld AI, "Memory Architecture Whitepaper," Inworld AI Inc., San Francisco, CA, 2025.
[59] NVIDIA Corporation, "NVIDIA ACE: Real-Time AI-Powered Game Characters," NVIDIA GTC Technical Session, Santa Clara, CA, 2025.
[60] J. Kim et al., "Emotion-Conditioned Neural TTS for Game NPC Voice Synthesis," in Proc. INTERSPEECH, Dublin, Ireland, 2024.
[61] T. Tulving, "Episodic Memory: From Mind to Brain," Annu. Rev. Psychol., vol. 53, pp. 1-25, 2002.
[62] J. Johnson, M. Douze, and H. Jégou, "Billion-Scale Similarity Search with GPUs," IEEE Trans. Big Data, vol. 7, no. 3, pp. 535-547, 2021.
[63] Inworld AI, "Memory Importance Scoring in Production NPC Systems," Blog.inworld.ai, 2025.
[64] Epic Games, "Privacy-Preserving NPC Memory: Technical Approach," Epic Developer Community Blog, 2025.
[65] D. Livingstone, "Turing's Test and Believable AI in Games," Comput. Entertain., vol. 4, no. 1, 2006.
[66] Ortony, G. L. Clore, and A. Collins, The Cognitive Structure of Emotions. Cambridge University Press, 1988.
[67] NVIDIA Corporation, "Emotional State Machine in NVIDIA ACE," NVIDIA Developer Documentation, Santa Clara, CA, 2025.
[68] K. Scherer, "Appraisal Theory," in Handbook of Cognition and Emotion, T. Dalgleish and M. Power, Eds. Wiley, 1999.
[69] Ubisoft, "NEO NPCs: The Next Generation of Game Characters," Ubisoft Research Blog, Paris, France, 2024.
[70] Ubisoft Research, "NEO NPC Player Testing Results: Believability and Engagement Metrics," Ubisoft Technical Report, Paris, France, 2025.
[71] J. Orkin, "Symbolic Behavior Representations for AI Character Behavior," in Game AI Pro 3, CRC Press, 2017.
[72] T. Williams et al., "AF-G: LLM-GOAP Integration Framework for Adaptive Game Agents," in Proc. IEEE CoG, 2025.
[73] P. Hart, N. Nilsson, and B. Raphael, "A Formal Basis for the Heuristic Determination of Minimum Cost Paths," IEEE Trans. Syst. Sci. Cybern., vol. 4, no. 2, pp. 100-107, 1968.
[74] Epic Games, "StateTree: Hierarchical State Machine for Unreal Engine 5," Epic Unreal Engine Docs, 2025.
[75] Epic Games, "LLM Integration with StateTree: Reference Architecture," Epic Developer Community, 2025.
[76] Khalifa et al., "Procedural Level Generation via Deep Learning: Evaluation Framework," in Proc. IEEE CoG, 2025.
[77] S. Sudhakaran et al., "Transformer-Based Minecraft World Generation," arXiv preprint arXiv:2501.12340, 2025.
[78] Epic Games, "Fortnite AI-Assisted Map Generation Pipeline," GDC Presentation, San Francisco, CA, 2025.
[79] N. Walton, "AI Dungeon and the Future of Interactive Fiction," Latitude Inc. Blog, 2025.
[80] Bethesda Game Studios, "Radiant Quest Augmentation with Language Models," GDC Research Session, 2025.
[81] M. Goodfellow et al., "Generative Adversarial Networks for Game Asset Production," arXiv preprint arXiv:1406.2661v4, 2014. [Production applications: 2025 game dev surveys]
[82] Adobe, "Substance 3D AI Tools: Studio Adoption Report," Adobe Creative Cloud Blog, 2025.
[83] SUNO AI, "SUNO Platform Technical Overview," SUNO AI Inc., Cambridge, MA, 2025.
[84] Variety, "SUNO AI, Udio Settle Record Label Lawsuits Over AI Training Data," Variety Media LLC, Los Angeles, CA, 2025.
[85] B. Mildenhall et al., "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis," Commun. ACM, vol. 65, no. 1, pp. 99-106, 2021.
[86] TripoSR, "TripoSR: Fast 3D Object Generation From Single Images," Tripo3D Inc., 2025.
[87] Game Developers Survey, "Text-to-3D in Production: Workflow Integration Study," GDC Vault, 2026.
[88] NVIDIA, "NVIDIA Omniverse AI Scene Generation Tools," NVIDIA Developer Documentation, 2025.
[89] D. Isla, "Halo 2 AI Using Goal-Oriented Action Planning," in Proc. Game Developers Conf., San Jose, CA, 2005.
[90] Inworld AI, "Tiered LLM Architecture for Production NPC Cost Optimization," Inworld Blog, 2025.
[91] C. Guo et al., "LoRA-Game: Fine-Tuning Llama-3 for Game NPC Dialogue," arXiv preprint arXiv:2505.08932, 2025.
[92] P. Christiano et al., "Deep Reinforcement Learning from Human Preferences," in Adv. Neural Inf. Process. Syst., 2017.
[93] J. Hinton et al., "Distilling the Knowledge in a Neural Network," arXiv preprint arXiv:1503.02531, 2015.
[94] Inworld AI, "Safety and Reliability in Production NPC Systems: 2025 Platform Report," Inworld AI Inc., 2025.
[95] Askell et al., "A General Language Assistant as a Laboratory for Alignment," arXiv preprint arXiv:2112.00861, 2021.
[96] M. Wooldridge and N. Jennings, "Intelligent Agents: Theory and Practice," Knowl. Eng. Rev., vol. 10, no. 2, pp. 115-152, 1995.
[97] R. Brooks, "A Robust Layered Control System for a Mobile Robot," IEEE J. Robot. Autom., vol. 2, no. 1, pp. 14-23, 1986.
[98] X. B. Peng et al., "AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control," ACM Trans. Graph., vol. 40, no. 4, 2021.
[99] T. Bansal et al., "Emergent Complexity via Multi-Agent Competition," in Proc. 6th ICLR, Vancouver, Canada, 2018.
[100] Electronic Arts, "AI-Driven Adaptive Difficulty in Sports Games," EA Research Blog, Redwood City, CA, 2025.
[101] K. Shaker, G. N. Yannakakis, and J. Togelius, "Towards Player-Driven Procedural Content Generation," in Proc. 7th Int. Conf. Found. Digit. Games, Raleigh, NC, 2012.
[102] R. Laban et al., "HRI Perceptions of Response Latency: A User Study," in Proc. ACM/IEEE HRI, 2022.
[103] T. Dao et al., "FlashAttention-2: Faster Attention with Better Parallelism," in Proc. 12th ICLR, 2024.
[104] Y. Leviathan, M. Kalman, and Y. Matias, "Fast Inference from Transformers via Speculative Decoding," in Proc. 40th ICML, 2023.
[105] NVIDIA Corporation, "TensorRT-LLM: Production LLM Inference Optimization," NVIDIA Developer Blog, 2025.
[106] G. Chen et al., "Response Pre-Generation Caching for Real-Time NPC Dialogue Systems," in Proc. IEEE CoG, 2025.
[107] NVIDIA Corporation, "Streaming Speech Synthesis for ACE NPCs," NVIDIA Technical Report, 2025.
[108] J. Gregory, Game Engine Architecture, 3rd ed. CRC Press, 2019.
[109] B. Schwab, AI Game Engine Programming, 2nd ed. Course Technology PTR, 2009
[110] NVIDIA Corporation, "RTX AI Toolkit: Local LLM Inference Benchmarks," NVIDIA Performance Report, Santa Clara, CA, 2025.
[111] Microsoft, "DirectML AI Inference APIs for DirectX 12," Microsoft Developer Blog, Redmond, WA, 2026.
[112] L. Evans, "Scalable AI for Open-World Games: Hierarchical NPC Budget Systems," in Game AI Pro 4, CRC Press, 2025.
[113] Epic Games, "GDC 2025: AI-Powered NPCs in Fortnite: Technical Architecture," GDC Vault, San Francisco, CA, 2025.
[114] Epic Games, "Adaptive Persona Architecture for Seasonal Narrative NPC Alignment," Epic Developer Community, 2025.
[115] Rockstar Games, "US Patent Application 20240321183: NPC Memory Compression and Dialogue Decay," USPTO, Washington, DC, 2024.
[116] Rockstar Games, "US Patent 11,942,101: Personality Drift Modeling for Autonomous Game Characters," USPTO, Washington, DC, 2024.
[117] Ubisoft Research, "NEO NPC Technical Architecture: Character, World, and Experience Engines," Ubisoft Developer Blog, Paris, France, 2024.
[118] J.-F. Dugas et al., "AI-Driven Dynamic Quest Generation in Open-World Games," in Proc. AIIDE, 2025.
[119] Ubisoft Research, "NEO NPC Living Quest Player Study," Ubisoft Internal Research Report (Disclosed Summary), Paris, France, 2025.
[120] NVIDIA Corporation, "NVIDIA-Inworld AI Technology Partnership Announcement," NVIDIA Press Release, Santa Clara, CA, 2024.
[121] Epic Games, "AI Tools in Unreal Engine 5: Production Suite Overview," Epic GDC 2025 Presentation.
[122] Epic Games, "UEFN AI Creative Tools: 2025 Year in Review," Epic Games Creator Blog, 2025.
[123] Liapis et al., "Sentient Sketchbook: Computer-Assisted Game Level Authoring," in Proc. 8th Int. Conf. Found. Digit. Games, 2013.
[124] SafeguardAI, "Behavioral Constraint Systems for Production Game AI," SafeguardAI Technical Report, 2025.
[125] R. Perez-Alonso et al., "Prompt Injection in NPC Systems: Detection and Mitigation," in Proc. GameSec, 2025.
[126] P. Lankoski and S. Bjork, "Gameplay Design Patterns for Believable Non-Player Characters," in Proc. DiGRA, 2007.
[127] H. Morales, "Controlled Imperfection: Designing for AI NPC Believability," in Proc. CHI Play, 2025.
[128] S. Yuan et al., "Accessibility of Conversational AI in Games for Players with Cognitive Disabilities," in Proc. ASSETS, 2025.
[129] KAI (Knowledge & AI Institute), "State of AI Ethics in Game Development 2026," Annual Survey Report, 2026.
[130] SAG-AFTRA, "2023 Memorandum of Agreement: Artificial Intelligence Provisions," SAG-AFTRA, Los Angeles, CA, 2023.
[131] SAG-AFTRA, "2025 Interactive Media Agreement: AI Voice Replication Terms," SAG-AFTRA, Los Angeles, CA, 2025.
[132] Zou et al., "Universal and Transferable Adversarial Attacks on Aligned Language Models," arXiv preprint arXiv:2307.15043, 2023.
[133] Y. Bai et al., "Constitutional AI: Harmlessness from AI Feedback," arXiv preprint arXiv:2212.08073, 2022.
[134] E. Bender et al., "On the Dangers of Stochastic Parrots," in Proc. ACM FAccT, Virtual, 2021.
[135] US Copyright Office, "Copyright and Artificial Intelligence: Policy Guidance 2025," US Copyright Office, Washington, DC, 2025.
[136] Bloomberg Law, "AI Training Data Litigation: 2025 Status Report," Bloomberg Law, New York, NY, 2025.
[137] European Parliament, "Artificial Intelligence Act (EU) 2024/1689," Official Journal of the European Union, 2024.
[138] Entertainment Software Rating Board (ESRB), "AI Content Disclosure Framework: Draft Guidelines," ESRB, New York, NY, 2026.
[139] Inworld AI, "Platform Analytics: Player Retention Correlates with NPC AI Quality," Inworld AI Blog, 2026.
[140] T. Zhang et al., "BERTScore: Evaluating Text Generation with BERT," in Proc. 8th ICLR, 2020.
[141] M. Chen et al., "GameDialogBench: Benchmark for Evaluating AI Dialogue in Game Contexts," arXiv preprint arXiv:2506.01123, 2025.
[142] P. Wang et al., "CharacterConsistencyEval: Automated Evaluation of NPC Personality Coherence," in Proc. EMNLP, 2025.
[143] G. N. Yannakakis and J. Togelius, "Experience-Driven Procedural Content Generation," IEEE Trans. Affect. Comput., vol. 2, no. 3, pp. 147-161, 2011.
[144] D. Isla, "The Future of Game AI: Hybrid Architectures and Generative Models," AIIDE Keynote, 2025.
[145] T. Significant Gravitas, "AutoGPT: Autonomous GPT-4 Agent," GitHub Repository, 2023.
[146] R. Twose, "Agentic NPCs and Game Balance: Open Problems," in Proc. IEEE CoG Workshop on Game AI, 2025.
[147] J. E. Laird, A. Newell, and P. S. Rosenbloom, "SOAR: An Architecture for General Intelligence," Artif. Intell., vol. 33, no. 1, pp. 1-64, 1987.
[148] J. R. Anderson, "ACT: A Simple Theory of Complex Cognition," Am. Psychol., vol. 51, no. 4, pp. 355-365, 1996.
[149] S. Franklin et al., "LIDA: A Cognitive Architecture Independent of Computational Substrate," in Proc. AGI Conf., 2013.
[150] Ng et al., "Persistent Memory in AI NPC Systems: Challenges and Solutions," in Proc. AIIDE, 2025.
[151] Mistral AI, "Mistral Large 2: 128K Context Technical Report," Mistral AI, Paris, France, 2025.
[152] N. F. Liu et al., "Lost in the Middle: How Language Models Use Long Contexts," Trans. Assoc. Comput. Linguist., vol. 12, pp. 157-173, 2024.
[153] H.-A. Park et al., "Scaling Generative Agent Simulations: Toward Thousand-Agent Social Worlds," in Proc. NeurIPS Workshop on Foundation Models, 2025.
[154] B. Liu et al., "Efficient Multi-Agent Simulation for Open-World Game Population Modeling," arXiv preprint arXiv:2506.11432, 2025.
[155] Zeng et al., "VLM Game Agents: Visual Perception for NPC World Understanding," in Proc. IEEE CoG, 2025.
[156] NVIDIA Research, "GPT-4V for Game-State-Aware Agent Control," NVIDIA Research Blog, 2025.
[157] AI Dungeon, "Platform Statistics and User Engagement Report 2026," Latitude Inc., Provo, UT, 2026.
