Pangjing Wu
Author directory2026
Datamart-Agent: LLM-Driven Game-Theoretic Agent for Data Marketplace Modeling
Pangjing Wu | Peter Q. Chen | Xiaodong Li | Wenqi Fan | Li Qing
Findings of the Association for Computational Linguistics: ACL 2026
Pangjing Wu | Peter Q. Chen | Xiaodong Li | Wenqi Fan | Li Qing
Findings of the Association for Computational Linguistics: ACL 2026
Data marketplaces analyze strategic data exchanges among users, platforms, and buyers. However, most existing studies model static equilibria and complete information, which limits their realism. In this work, we study whether large language model (LLM)-driven agents can make equilibrium-consistent decisions in analytically tractable data marketplaces with evolving and incomplete-information. Specifically, we introduce EvoDM, an agent-based modeling framework that extends the static data marketplace to dynamic and incomplete-information settings while providing tractable equilibrium benchmarks for evaluating agent decisions. Building upon EvoDM, we propose Datamart-Agent, an LLM-driven game-theoretic agent that improves equilibrium-consistent decision execution through dynamic game tree memory and mechanism-guided reflection, without requiring parameter updates. Experiments demonstrate that Datamart-Agent closely matches equilibrium-consistent decision-making, achieving the lowest utility gap and over 20% higher Pass@𝜖 than strong baselines. After validating its effectiveness, we employ EvoDM with Datamart-Agent to analyze competition and regulation in assumption-relaxed settings where closed-form ground truth is unavailable, providing exploratory simulation-based insights into market dynamics and regulatory effects.
“Do You Truly Love Me?” Benchmarking LLM Capability on Hierarchical Pragmatic Tactic Conversations
Peter Q. Chen | Pangjing Wu | Zhuohang Jiang | Xiaodong Li | Peter H. F. Ng
Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
Peter Q. Chen | Pangjing Wu | Zhuohang Jiang | Xiaodong Li | Peter H. F. Ng
Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
Pragmatic reasoning, including inferring intent, manipulation, and hidden meaning in dialogue, is essential for trustworthy language understanding yet remains underexplored in current large language models (LLMs). We introduce HIPO, a new benchmark designed to assess Hierarchical Pragmatic Tactic in multi-turn Conversations. Inspired by linguistic pragmatics, HIPO operationalizes a hierarchical annotation schema for multi-turn conversations and labels each utterance along one dialogue-level goal and three utterance-level dimensions: communicative intention (why), veracity strategy (how), and illocutionary act (what). To ensure reliable supervision, we design a structured generation pipeline to generate high-quality synthetic dialogues. The benchmark comprises over 4,088 benchmarking utterances with 6,350 contextual utterances across 1,131 dialogues drawn from 31 real-world-inspired scenarios. Benchmarking 22 state-of-the-art LLMs reveals a striking gap: while models achieve higher performance at recognizing surface speech acts(83.3% accuracy), they perform poorly on detecting veracity strategies(32.2%) and speaker intentions(45.1%). These results suggest current LLMs remain challenged by the pragmatic inferences operationalized in HIPO, particularly those involving information manipulation and intended effects beyond literal content. We further show that supervised fine-tuning on additional synthetic data generated by our pipeline can improve performance on HIPO, suggesting a practical path for targeted capability improvement within this benchmark setting.
Search