Combining LLM, Non-Monotonic Logical Reasoning, and Human-In-the-loop Feedback in an Assistive AI Agent
The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.
Abstract
Large Language Models (LLMs) are considered state of the art for many tasks in robotics and AI. At the same time, there is increasing evidence of their critical limitations such as generating arbitrary responses in new situations, inability to support rapid incremental updates based on limited examples, and opacity. Toward addressing these limitations, our architecture leverages the complementary strengths of LLMs and knowledge-based reasoning. Specifically, the architecture enables an AI agent assisting a human to use an LLM to provide generic abstract predictions of upcoming tasks. The agent also reasons with domain-specific knowledge, recent history of interactions with the human, and semantic databases to: (a) provide contextual prompts to the LLM; and (b) compute a plan of concrete actions that jointly implements the current task and prepares for the anticipated task, replanning as needed. Furthermore, the agent solicits and uses high-level human feedback based on need and availability to incrementally revise the domain-specific knowledge and interactions with the LLM. We ground and evaluate our architecture’s abilities in the realistic VirtualHome simulation environment, demonstrating a substantial performance improvement compared with just using an LLM or an LLM and logical reasoner. Project website: https://brianej.github.io/igfmrdskaa.github.io/
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Fu et al. extended the concept of combining language models with human judgment by incorporating non-monotonic logical reasoning, thereby creating assistive AI agents with more robust reasoning capabilities under uncertain situations [49].”
From Human-in-the-Loop Artificial Intelligence: A Systematic Review of Concepts, Methods, and Applications (2026) · cites this paper as [49]
Citations
Cites
WordNet: a lexical database for English
Miller, G. A. (1995) · Communications of the ACM · doi:10.1145/219717.219748 · Cited by 14,015
Power to the People: The Role of Humans in Interactive Machine Learning
Amershi, S., Cakmak, M., Knox, W. B., & Kulesza, T. (2014) · AI Magazine · doi:10.1609/aimag.v35i4.2513 · Cited by 1,084
LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models
Song, C. H., Sadler, B. M., Wu, J., Chao, W.-L., Washington, C., & Su, Y. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00280 · Cited by 435 · cited 1×
“In complex domains, human feedback plays a vital role in ensuring reliable and efficient operation in the context of planning, learning, and other tasks [23]–[25].”
VirtualHome: Simulating Household Activities Via Programs
Puig, X., Ra, K., Boben, M., Li, J., Wang, T., Fidler, S., & Torralba, A. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00886 · Cited by 378
Gelfond, M., & Kahl, Y. (2014) · doi:10.1017/cbo9781139342124 · Cited by 253
Loss of plasticity in deep continual learning
Dohare, S., Hernandez-Garcia, J. F., Lan, Q., Rahman, P., Mahmood, A. R., & Sutton, R. S. (2024) · Nature · doi:10.1038/s41586-024-07711-7 · Cited by 172 · cited 1×
“In particular, they are based on a narrow set of representations and update processes; make arbitrary decisions in truly novel situations; do not support rapid, incremental updates from limited examples; and are opaque [6], [8], [9].”
Open-vocabulary Queryable Scene Representations for Real World Planning
Chen, B., Xia, F., Ichter, B., Rao, K., Gopalakrishnan, K., Ryoo, M. S., Stone, A., & Kappler, D. (2023) · 2023 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra48891.2023.10161534 · Cited by 143
Robot learning on the job: Human-in-the-loop autonomy and learning during deployment
Liu, H., Nasiriany, S., Zhang, L., Bao, Z., & Zhu, Y. (2024) · The International Journal of Robotics Research · doi:10.1177/02783649241273901 · Cited by 75
HDDL: An Extension to PDDL for Expressing Hierarchical Planning Problems
Höller, D., Behnke, G., Bercher, P., Biundo, S., Fiorino, H., Pellier, D., & Alford, R. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i06.6542 · Cited by 72
Towards Answer Set Programming with Sorts
Balai, E., Gelfond, M., & Zhang, Y. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-40564-8_14 · Cited by 53
Research works printed but not linked: 16
- Doshi, Scaling Cross-Embodied Learning: One Policy for Manipulation, Navigation, Locomotion, and Aviation
- Huang, Language models as zero-shot planners: Extracting actionable knowledge for embodied agents
- Liu, 2023, LLLM+P: Empowering Large Language Models with Optimal Planning Proficiency
- Zhang, Building Cooperative Embodied Agents Modularly with Large Language Models
- Zhao, Large Language Models as Commonsense Knowledge for Large-Scale Task Planning
- Kambhampati, 2024, Llms can’t plan, but can help planning in llm-modulo frameworks
- Silver, Pddl planning with pretrained large language models
- Guan, 2023, Lever-aging pre-trained large language models to construct and utilize world models for model-based task planning, Advances in Neural Information Processing Systems, 36
- Schick, 2023, Toolformer: Language Models Can Teach Themselves to Use Tools, Advances in Neural Information Processing Systems
- 2024, GPT-4o System Card, abs/2410.21276 · arXiv
- Singh, 2024, Anticipate & col-lab: Data-driven task anticipation and knowledge-driven planning for human-robot collaboration
- Driess, 2023, Palm-e: An embodied multimodal language model
- Gao, 2024, Retrieval-augmented generation for large language models: A survey
- Shaik, 2024, An overview and discussion on using large language models for implementation generation of solutions to open-ended problems
- Shinn, 2024, Reflexion: Language agents with verbal reinforcement learning, Advances in Neural Information Processing Systems, 36
- Kasaei, 2024, Vital: Visual teleoperation to enhance robot learning through human-in-the-loop corrections
Cited by
Lazaros, K., Vrahatis, A. G., & Kotsiantis, S. (2026) · Entropy · doi:10.3390/e28040377 · Cited by 50
Human-AI Agentic Approach for Industrial Software Systems Design
Pegado, A. M., Camarinha-Matos, L. M., & Rocha, A. D. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-28939-1_1
Toward Reliable Agentic AI for Software Architectural Analysis
Liljas, V. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-39143-8_4
Metadata
- Authors
- Fu, T., Jauw, B., & Sridharan, M.
- Affiliations
- University of Edinburgh (GB)
- ORCID
- Sridharan, Mohan
- Venue
- 2025 34th IEEE International Conference on Robot and Human Interactive Communication (RO-MAN)
- Publisher
- IEEE
- Published
- 2025-08-25
- Pages
- 322-329
- ISSN
- 1944-9437 (ISSN-L)
- OpenAlex
- W4415821765
- Cited by
- 3
- References
- 18
- Citations by year
- 2026: 3
- FWCI
- 2.86
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE RO-MAN (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › AI-based Problem Solving and Planning
- Physical Sciences › Computer Science › Artificial Intelligence › Logic, Reasoning, and Knowledge
- Physical Sciences › Engineering › Control and Systems Engineering › Robot Manipulation and Learning
- Keywords
- large language models, nonmonotonic logic, human-in-the-loop feedback, human-robot interaction, task prediction, action planning, replanning, knowledge revision, semantic databases, knowledge reasoning
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Combining LLM, Non-Monotonic Logical Reasoning, and Human-In-the-loop Feedback in an Assistive AI Agent. https://app.cochat.ai/papers/10.1109/ro-man63969.2025.11217618
- The paper
- Fu, T., Jauw, B., & Sridharan, M. (2025). Combining LLM, Non-Monotonic Logical Reasoning, and Human-In-the-loop Feedback in an Assistive AI Agent. 2025 34th IEEE International Conference on Robot and Human Interactive Communication (RO-MAN). https://doi.org/10.1109/ro-man63969.2025.11217618
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.