app/src/content/bibliography.bib

% Agent State Graph - Related Work Bibliography

% === Our Previous Work ===

@inproceedings{wu2025agentgraph,
  title={{AgentGraph}: Trace-to-Graph Platform for Interactive Analysis and Robustness Testing in Agentic {AI} Systems},
  author={Wu, Zekun and Cho, Seonglae and Munoz, Cristian and King, Theo and Mohammed, Umar and Kazim, Emre and Perez-Ortiz, Maria and Bulathwela, Sahan and Koshiyama, Adriano},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  year={2025},
  note={Platform paper. LLM-driven FSM construction from agent traces + perturbation testing.}
}

% === Most Related: Trace-based Workflow/State Extraction ===

@article{wang2024agent_workflow_memory,
  title={Agent Workflow Memory},
  author={Wang, Zora Zhiruo and Mao, Jiayuan and Fried, Daniel and Neubig, Graham},
  journal={arXiv preprint arXiv:2409.07429},
  year={2024},
  note={Extracts reusable workflow patterns from successful agent traces. Linear workflows, not FSMs.}
}

@article{xia2025experience_to_strategy,
  title={From Experience to Strategy: Empowering {LLM} Agents with Trainable Graph Memory},
  author={Xia, Siyu and Xu, Zekun and Chai, Jiajun and Fan, Wentian and Song, Yan and Wang, Xiaohan and Yin, Guojun and Lin, Wei and Zhang, Haifeng and Wang, Jun},
  journal={arXiv preprint arXiv:2511.07800},
  year={2025},
  note={Abstracts trajectories into state machine + meta-cognition for RL training.}
}

@article{zhao2024expel,
  title={{ExpeL}: {LLM} Agents Are Experiential Learners},
  author={Zhao, Andrew and Huang, Daniel and Xu, Quentin and Lin, Matthieu and Liu, Yong-Jin and Huang, Gao},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  year={2024},
  note={Extracts insights from traces as natural language rules, not structured graphs.}
}

% === FSM + Agent Systems ===

@inproceedings{chen2025metaagent,
  title={{MetaAgent}: Automatically Constructing Multi-Agent Systems Based on Finite State Machines},
  author={Chen, Zuoxiang and Wu, Yiwen and Li, Zhenhailong and Ji, Heng},
  booktitle={Proceedings of the International Conference on Machine Learning (ICML)},
  year={2025},
  note={Top-down FSM generation from task descriptions. Not trace-based.}
}

@article{wu2024stateflow,
  title={{StateFlow}: Enhancing {LLM} Task-Solving through State-Driven Workflows},
  author={Wu, Yiran and Yue, Tianwei and Zhang, Shaokun and Wang, Chi and Wu, Qingyun},
  journal={arXiv preprint arXiv:2403.11322},
  year={2024},
  note={Manually defined FSMs for agent orchestration in AutoGen.}
}

@article{ali2024flowfsm,
  title={An Agentic Flow for Finite State Machine Extraction using Prompt Chaining},
  author={Ali, Youssef Maklad and others},
  journal={arXiv preprint arXiv:2507.11222},
  year={2025},
  note={FSM extraction from RFC documents, not execution traces.}
}

@article{sun2024dfa_llm,
  title={Chatbot Meets Pipeline: Augment Large Language Model with Definite Finite Automaton},
  author={Wang, Yiyou and others},
  journal={arXiv preprint arXiv:2402.04411},
  year={2024},
  note={DFA structure learned from dialogues to constrain LLM output.}
}

@inproceedings{umili2024deepdfa,
  title={{DeepDFA}: Automata Learning through Neural Probabilistic Relaxations},
  author={Umili, Elena and Capobianco, Roberto},
  booktitle={Proceedings of the European Conference on Artificial Intelligence (ECAI)},
  year={2024},
  note={Differentiable DFA extraction via gradient-based optimization on probabilistic relaxations.}
}

@inproceedings{wang2026agentspec,
  title={{AgentSpec}: Customizable Runtime Enforcement for Safe and Reliable {LLM} Agents},
  author={Wang, Haoyu and Poskitt, Christopher M. and Sun, Jun},
  booktitle={Proceedings of the International Conference on Software Engineering (ICSE)},
  year={2026},
  note={DSL for specifying and enforcing runtime constraints on LLM agents.}
}

% === Workflow Generation / Search ===

@inproceedings{zhang2025aflow,
  title={{AFlow}: Automating Agentic Workflow Generation},
  author={Zhang, Jiayi and Xiang, Jinyu and Yu, Zhaoyang and Teng, Fengwei and Chen, Xiaobin and Lu, Jiaqi and Zhong, Mingchen and Zhang, Ming and Wang, Yuqiang and Li, Qian and Hong, Hongxia},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2025},
  note={ICLR 2025 Oral. MCTS over code-represented workflow space.}
}

@article{hu2024adas,
  title={{ADAS}: Automated Design of Agentic Systems},
  author={Hu, Shunyu and Lu, Xinyuan and others},
  journal={arXiv preprint arXiv:2408.08435},
  year={2024},
  note={Meta-agent iteratively designs agent architectures.}
}

@article{jin2024flowmind,
  title={{FlowMind}: Automatic Workflow Generation with {LLMs}},
  author={Jin, Zhen and others},
  journal={arXiv preprint},
  year={2024},
  note={Forward workflow generation from descriptions, not reverse from traces.}
}

% === Process Mining (Foundational) ===

@book{vanderaalst2016process,
  title={Process Mining: Data Science in Action},
  author={van der Aalst, Wil M. P.},
  publisher={Springer},
  year={2016},
  edition={2},
  note={Foundational text. Alpha algorithm, conformance checking, Petri nets.}
}

@article{grohs2024process_mining_llm,
  title={Evaluating Large Language Models in Process Mining: Capabilities, Benchmarks, Evaluation Strategies, and Future Challenges},
  author={Grohs, Alessandro and others},
  journal={arXiv preprint arXiv:2403.06749},
  year={2024},
  note={Survey of LLM capabilities on process mining tasks.}
}

@inproceedings{chen2024skill_process_mining,
  title={Skill Learning Using Process Mining for Large Language Model Plan Generation},
  author={Chen, Zijian and others},
  booktitle={International Conference on Process Mining (ICPM)},
  year={2024},
  note={Process discovery for LLM skill acquisition and retrieval.}
}

% === Automata Learning (Foundational) ===

@article{angluin1987learning,
  title={Learning Regular Sets from Queries and Counterexamples},
  author={Angluin, Dana},
  journal={Information and Computation},
  volume={75},
  number={2},
  pages={87--106},
  year={1987},
  note={L* algorithm. Active automata learning foundation.}
}

@article{vaandrager2017model,
  title={Model Learning},
  author={Vaandrager, Frits},
  journal={Communications of the ACM},
  volume={60},
  number={2},
  pages={86--95},
  year={2017},
  note={Survey of automata learning / model learning.}
}

@article{weiss2018extracting,
  title={Extracting Automata from Recurrent Neural Networks Using Queries and Counterexamples},
  author={Weiss, Gail and Goldberg, Yoav and Yahav, Eran},
  booktitle={Proceedings of the International Conference on Machine Learning (ICML)},
  year={2018},
  note={FSM extraction from neural networks via state merging.}
}

% === Datasets (All Peer-Reviewed) ===

@inproceedings{liu2024agentbench,
  title={{AgentBench}: Evaluating {LLMs} as Agents},
  author={Liu, Xiao and Yu, Hao and Zhang, Hanchen and Xu, Yifan and Lei, Xuanyu and Lai, Hanyu and Gu, Yu and Ding, Hangliang and Men, Kaiwen and Yang, Kejuan and others},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024},
  note={8-environment agent benchmark. HuggingFace: THUDM/AgentBench.}
}

@inproceedings{jimenez2024swebench,
  title={{SWE-bench}: Can Language Models Resolve Real-World {GitHub} Issues?},
  author={Jimenez, Carlos E. and Yang, John and Wettig, Alexander and Yao, Shunyu and Pei, Kexin and Press, Ofir and Narasimhan, Karthik},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024},
  note={2,294 real-world coding tasks. HuggingFace: princeton-nlp/SWE-bench.}
}

@inproceedings{yang2025swesmith,
  title={{SWE-smith}: Scaling Data for Software Engineering Agents},
  author={Yang, John and Lieret, Kilian and Jimenez, Carlos E. and Wettig, Alexander and Khandpur, Kabir and Zhang, Yanzhe and Hui, Binyuan and Press, Ofir and Schmidt, Ludwig and Yang, Diyi},
  booktitle={Proceedings of the Annual Conference on Neural Information Processing Systems (NeurIPS), Datasets \& Benchmarks Track},
  year={2025},
  note={Spotlight. 50K SWE task instances from 128 repos. arXiv: 2504.21798. GitHub: SWE-bench/SWE-smith.}
}

@inproceedings{zhou2024webarena,
  title={{WebArena}: A Realistic Web Environment for Building Autonomous Agents},
  author={Zhou, Shuyan and Xu, Frank F. and Zhu, Hao and Zhou, Xuhui and Lo, Robert and Sridhar, Abishek and Cheng, Xianyi and Bisk, Yonatan and Fried, Daniel and Alon, Uri and Neubig, Graham},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024},
  note={812 web tasks across 4 domains. GitHub: web-arena-x/webarena.}
}

@inproceedings{hou2025agenttrek,
  title={{AgentTrek}: Agent Trajectory Synthesis via Guiding Replay with Web Tutorials},
  author={Hou, Yiheng and others},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2025},
  note={Synthesized web agent trajectories from tutorials. HuggingFace: SkyRain25/AgentTrek.}
}

@article{yang2025whoandwhen,
  title={Which Agent Causes Task Failures and When?},
  author={Yang, Kevin and others},
  journal={arXiv preprint arXiv:2505.00212},
  year={2025},
  note={184 annotated multi-agent failure traces. HuggingFace: Kevin355/Who\_and\_When.}
}

@inproceedings{zhang2026agentracer,
  title={{AgenTracer}: Who Is Inducing Failure in the {LLM} Agentic Systems?},
  author={Zhang, Yu and others},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2026},
  note={2,500+ annotated trajectory-error pairs. HuggingFace: AgenTracer/TracerTraj.}
}

% === Agent Frameworks (Trace Sources) ===

@article{wu2023autogen,
  title={{AutoGen}: Enabling Next-Gen {LLM} Applications via Multi-Agent Conversation},
  author={Wu, Qingyun and Bansal, Gagan and Zhang, Jieyu and Wu, Yiran and Li, Beibin and Zhu, Erkang and Jiang, Li and Zhang, Xiaoyun and Zhang, Shaokun and Liu, Jiale and others},
  journal={arXiv preprint arXiv:2308.08155},
  year={2023}
}

@article{hong2023metagpt,
  title={{MetaGPT}: Meta Programming for A Multi-Agent Collaborative Framework},
  author={Hong, Sirui and Zhuge, Mingchen and Chen, Jonathan and Zheng, Xiawu and Cheng, Yuheng and Zhang, Ceyao and Wang, Jinlin and Wang, Zili and Yau, Steven Ka Shing and Lin, Zijuan and others},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024}
}

% === Trace Analysis / Agent Debugging ===

@article{cheng2024trace_autodiff,
  title={Trace is the New {AutoDiff}: Unlocking Efficient Optimization of Computational Workflows},
  author={Cheng, Ching-An and Nie, Allen and Swaminathan, Adith},
  journal={arXiv preprint arXiv:2406.16218},
  year={2024},
  note={Traces as computation graphs for optimization, not structural extraction.}
}

@article{ma2024agentboard,
  title={{AgentBoard}: An Analytical Evaluation Board of Multi-turn {LLM} Agents},
  author={Ma, Chang and others},
  journal={arXiv preprint arXiv:2401.13178},
  year={2024},
  note={Fine-grained progress analysis with predefined milestones.}
}

@article{wang2023voyager,
  title={{Voyager}: An Open-Ended Embodied Agent with Large Language Models},
  author={Wang, Guanzhi and Xie, Yuqi and Jiang, Yunfan and Mandlekar, Ajay and Xiao, Chaowei and Zhu, Yuke and Fan, Linxi and Anandkumar, Anima},
  journal={arXiv preprint arXiv:2305.16291},
  year={2023}
}

% === Agent Safety / Red-teaming (for Path A framing) ===

@article{ruan2024toolemu,
  title={Identifying the Risks of {LM} Agents with an {LM}-Emulated Sandbox},
  author={Ruan, Yangjun and Dong, Honghua and Wang, Andrew and Pitis, Silviu and Zhou, Yongchao and Ba, Jimmy and Dubois, Yann and Maddison, Chris J and Hashimoto, Tatsunori},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024},
  note={Risk identification in agent systems via emulated environments.}
}

% === Additional Datasets (NeurIPS Extension) ===

@article{yang2024sweagent,
  title={{SWE}-agent: Agent-Computer Interfaces Enable Automated Software Engineering},
  author={Yang, John and Jimenez, Carlos E. and Wettig, Alexander and Lieret, Kilian and Yao, Shunyu and Narasimhan, Karthik and Press, Ofir},
  journal={Advances in Neural Information Processing Systems (NeurIPS)},
  year={2024},
  note={Agent-computer interface for SWE-bench. HuggingFace: nebius/SWE-agent-trajectories.}
}

@inproceedings{deng2024mind2web,
  title={{Mind2Web}: Towards a Generalist Agent for the Web},
  author={Deng, Xiang and Gu, Yu and Zheng, Boyuan and Chen, Shijie and Stevens, Sam and Wang, Boshi and Sun, Huan and Su, Yu},
  booktitle={Advances in Neural Information Processing Systems (NeurIPS)},
  year={2023},
  note={2,350 web tasks from 137 websites. HuggingFace: osunlp/Mind2Web.}
}

@article{yao2024taubench,
  title={$\tau$-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains},
  author={Yao, Shunyu and Narasimhan, Karthik and others},
  journal={arXiv preprint arXiv:2406.12045},
  year={2024},
  note={Airline (200) + retail (460) customer service agent traces. GitHub: sierra-research/tau-bench.}
}

@article{barres2025tau2,
  title={$\tau^2$-bench: Evaluating Agents in Dynamic Multi-Turn Interactions},
  author={Jiang, Zihao and Zhang, Jingyu and Yao, Shunyu and Narasimhan, Karthik and others},
  journal={arXiv preprint arXiv:2506.07982},
  year={2025},
  note={Multi-model evaluation: 4 LLMs $\times$ 3 domains (airline, retail, telecom). HuggingFace: HuggingFaceH4/tau2-bench-data.}
}

@inproceedings{qin2024toolbench,
  title={{ToolLLM}: Facilitating Large Language Models to Master 16000+ Real-world {APIs}},
  author={Qin, Yujia and Liang, Shihao and Ye, Yining and Zhu, Kunlun and Yan, Lan and Lu, Yaxi and Lin, Yankai and Cong, Xin and Tang, Xiangru and Qian, Bill and others},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2024},
  note={Spotlight. 16,000+ APIs. HuggingFace: OpenBMB/ToolBench.}
}

% === Automata Learning (Additional) ===

@inproceedings{oncina1992rpni,
  title={Inferring Regular Languages in Polynomial Updated Time},
  author={Oncina, Jos{\'e} and Garc{\'\i}a, Pedro},
  booktitle={Pattern Recognition and Image Analysis},
  pages={49--61},
  year={1992},
  publisher={World Scientific},
  note={RPNI algorithm for grammatical inference from positive examples.}
}

@inproceedings{lang1998edsm,
  title={Results of the {Abbadingo} One {DFA} Learning Competition and a New Evidence-Driven State Merging Algorithm},
  author={Lang, Kevin J. and Pearlmutter, Barak A. and Price, Rodney A.},
  booktitle={International Colloquium on Grammatical Inference (ICGI)},
  pages={1--12},
  year={1998},
  note={EDSM: Evidence-Driven State Merging for DFA learning.}
}

@article{muskardin2022aalpy,
  title={{AALpy}: An Active Automata Learning Library},
  author={Mu{\v{s}}kardin, Edi and Aichernig, Bernhard K. and Pill, Ingo and Pferscher, Andrea and Tappler, Martin},
  journal={Innovations in Systems and Software Engineering},
  volume={18},
  pages={417--426},
  year={2022},
  note={Python library for active and passive automata learning.}
}

% === Agent Safety / Benchmarks ===

@inproceedings{zhang2025asb,
  title={Agent Security Bench ({ASB}): Formalizing and Benchmarking Attacks and Defenses in {LLM}-based Agents},
  author={Zhang, Hanrong and Huang, Jingyuan and Mei, Kai and Yao, Yifei and Wang, Zhenting and Zhan, Chenlu and Wang, Hongwei and Zhang, Yongfeng},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)},
  year={2025},
  note={Formalizes adversarial attacks and defenses for LLM agents across 10 scenarios.}
}

@article{berti2024processmining,
  title={Re-Thinking Process Mining in the {AI}-Based Agents Era},
  author={Berti, Alessandro and Maatallah, Mayssa and Jessen, Urszula and Sroka, Michal and Ghannouchi, Sonia Ayachi},
  journal={arXiv preprint arXiv:2408.07720},
  year={2024},
  note={Proposes agent workflow paradigm to integrate process mining with LLM agents.}
}

@article{gold1967language,
  title={Language Identification in the Limit},
  author={Gold, E. Mark},
  journal={Information and Control},
  volume={10},
  number={5},
  pages={447--474},
  year={1967},
  note={Shows regular languages not identifiable from positive examples alone in the limit.}
}

@article{delaHiguera2010grammatical,
  title={Grammatical Inference: Learning Automata and Grammars},
  author={de la Higuera, Colin},
  publisher={Cambridge University Press},
  year={2010},
  note={Comprehensive text on automata learning.}
}

% === Graph Analysis (Neumann Series / Reachability) ===

@book{horn2012matrix,
  title={Matrix Analysis},
  author={Horn, Roger A. and Johnson, Charles R.},
  publisher={Cambridge University Press},
  year={2012},
  edition={2},
  note={Neumann series $(I-A)^{-1}$ for reachability/influence analysis.}
}

% === Theory References ===

@book{hopcroft2006automata,
  title={Introduction to Automata Theory, Languages, and Computation},
  author={Hopcroft, John E. and Motwani, Rajeev and Ullman, Jeffrey D.},
  publisher={Pearson},
  year={2006},
  edition={3},
  note={Myhill-Nerode theorem, DFA minimization.}
}

@book{motwani1995randomized,
  title={Randomized Algorithms},
  author={Motwani, Rajeev and Raghavan, Prabhakar},
  publisher={Cambridge University Press},
  year={1995},
  note={Coupon collector bounds.}
}

@inproceedings{carrasco1994alergia,
  title={Learning Stochastic Regular Grammars by Means of a State Merging Method},
  author={Carrasco, Rafael C. and Oncina, Jos{\'e}},
  booktitle={International Colloquium on Grammatical Inference},
  pages={139--152},
  year={1994},
  publisher={Springer},
  note={Alergia: learns PDFAs from positive examples using Hoeffding bound for state merging.}
}

@article{rabiner1989hmm,
  title={A Tutorial on Hidden {Markov} Models and Selected Applications in Speech Recognition},
  author={Rabiner, Lawrence R.},
  journal={Proceedings of the IEEE},
  volume={77},
  number={2},
  pages={257--286},
  year={1989},
  note={Classic HMM tutorial. Baum-Welch EM for parameter estimation.}
}

@inproceedings{lu2024guiodyssey,
  title={{GUI-Odyssey}: A Comprehensive Dataset for Cross-App {GUI} Navigation on Mobile Devices},
  author={Lu, Quanfeng and Zhao, Wenqi and Jia, Jiaming and Ren, Kanzhi and Lu, Kaiyi and Han, Jianbo and Chen, Yongxin and Zheng, Jingtao and Zhang, Zixuan and Ding, Liang},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2025},
  note={7,735 cross-app mobile GUI agent traces, 6 action types, 201 apps.}
}

@article{wang2025opencua,
  title={{OpenCUA}: Open Foundations for Computer-Use Agents},
  author={Wang, Xinyuan and Wang, Bowen and Lu, Dunjie and Yang, Junlin and Xie, Tianbao and others},
  journal={arXiv preprint arXiv:2508.09123},
  year={2025},
  note={22.6K desktop computer-use agent trajectories (Ubuntu/Windows/macOS), 11 action types. HuggingFace: xlangai/AgentNet.}
}


@inproceedings{xie2024osworld,
title={{OSW}orld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments},
author={Tianbao Xie and Danyang Zhang and Jixuan Chen and Xiaochuan Li and Siheng Zhao and Ruisheng Cao and Toh Jing Hua and Zhoujun Cheng and Dongchan Shin and Fangyu Lei and Yitao Liu and Yiheng Xu and Shuyan Zhou and Silvio Savarese and Caiming Xiong and Victor Zhong and Tao Yu},
booktitle={The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track},
year={2024},
url={https://openreview.net/forum?id=tN61DTr4Ed},
  note={369 real computer tasks across Ubuntu/Windows/macOS.}
}

@inproceedings{daciuk2000incremental,
  title={Incremental Construction of Minimal Acyclic Finite-State Automata},
  author={Daciuk, Jan and Mihov, Stoyan and Watson, Bruce W. and Watson, Richard E.},
  journal={Computational Linguistics},
  volume={26},
  number={1},
  pages={3--16},
  year={2000},
  note={Acyclic DFA minimisation via suffix sharing; foundational for the prefix-tree merging we apply to agent traces.}
}