eprintid: 10194858 rev_number: 7 eprint_status: archive userid: 699 dir: disk0/10/19/48/58 datestamp: 2024-07-19 13:34:12 lastmod: 2024-07-19 13:34:12 status_changed: 2024-07-19 13:34:12 type: proceedings_section metadata_visibility: show sword_depositor: 699 creators_name: Fang, M creators_name: Deng, S creators_name: Zhang, Y creators_name: Shi, Z creators_name: Chen, L creators_name: Pechenizkiy, M creators_name: Wang, J title: Large Language Models Are Neurosymbolic Reasoners ispublished: pub divisions: UCL divisions: B04 divisions: F48 note: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. abstract: A wide range of real-world applications is characterized by their symbolic nature, necessitating a strong capability for symbolic reasoning. This paper investigates the potential application of Large Language Models (LLMs) as symbolic reasoners. We focus on text-based games, significant benchmarks for agents with natural language capabilities, particularly in symbolic tasks like math, map reading, sorting, and applying common sense in text-based worlds. To facilitate these agents, we propose an LLM agent designed to tackle symbolic challenges and achieve in-game objectives. We begin by initializing the LLM agent and informing it of its role. The agent then receives observations and a set of valid actions from the text-based games, along with a specific symbolic module. With these inputs, the LLM agent chooses an action and interacts with the game environments. Our experimental results demonstrate that our method significantly enhances the capability of LLMs as automated agents for symbolic reasoning, and our LLM agent is effective in text-based games involving symbolic tasks, achieving an average performance of 88% across all tasks. date: 2024-03-25 date_type: published publisher: Association for the Advancement of Artificial Intelligence (AAAI) official_url: http://dx.doi.org/10.1609/aaai.v38i16.29754 oa_status: green full_text_type: other language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 2268264 doi: 10.1609/aaai.v38i16.29754 lyricists_name: Wang, Jun lyricists_id: JWANG00 actors_name: Wang, Jun actors_id: JWANG00 actors_role: owner full_text_status: public pres_type: paper publication: Proceedings of the AAAI Conference on Artificial Intelligence volume: 38 number: 16 pagerange: 17985-17993 event_title: The 38th Annual AAAI Conference on Artificial Intelligence issn: 2159-5399 book_title: Proceedings of the AAAI Conference on Artificial Intelligence citation: Fang, M; Deng, S; Zhang, Y; Shi, Z; Chen, L; Pechenizkiy, M; Wang, J; (2024) Large Language Models Are Neurosymbolic Reasoners. In: Proceedings of the AAAI Conference on Artificial Intelligence. (pp. pp. 17985-17993). Association for the Advancement of Artificial Intelligence (AAAI) Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10194858/1/2401.09334v1.pdf