# Evaluating Few-Shot Temporal Reasoning of LLMs for Human Activity Prediction in Smart Environments

arXiv: https://arxiv.org/abs/2602.11176 · HTML: https://mlsanity.com/paper/2602.11176/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 16 of its 32 references that are in the map.
- First posted on arXiv 20 Jan 2026.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)

## Builds on (most cited first, 16 of 16)

- [OpenCity: A Scalable Platform to Simulate Urban Activities with Massive LLM Agents](https://mlsanity.com/paper/2410.21286/) (2024)
- [CitySim: Modeling Urban Behaviors and City Dynamics with Large-Scale LLM-Driven Agent Simulation](https://mlsanity.com/paper/2506.21805/) (2025)
- [Looking Ahead: Anticipating Pedestrians Crossing with Future Frames Prediction](https://mlsanity.com/paper/1910.09077/) (2019)
- [Pedestrian Models for Autonomous Driving Part II: High-Level Models of Human Behavior](https://mlsanity.com/paper/2003.11959/) (2020)
- [A Survey of Human Activity Recognition in Smart Homes Based on IoT Sensors Algorithms: Taxonomies, Challenges, and Opportunities with Deep Learning](https://mlsanity.com/paper/2111.04418/) (2021)
- [Pedestrian Behavior Prediction for Automated Driving: Requirements, Metrics, and Relevant Features](https://mlsanity.com/paper/2012.08418/) (2020)
- [Smart Agent-Based Modeling: On the Use of Large Language Models in Computer Simulations](https://mlsanity.com/paper/2311.06330/) (2023)
- [GPT-4V Takes the Wheel: Promises and Challenges for Pedestrian Behavior Prediction](https://mlsanity.com/paper/2311.14786/) (2023)
- [From Text to Space: Mapping Abstract Spatial Models in LLMs during a Grid-World Navigation Task](https://mlsanity.com/paper/2502.16690/) (2025)
- [Combining Public Human Activity Recognition Datasets to Mitigate Labeled Data Scarcity](https://mlsanity.com/paper/2306.13735/) (2023)
- [SpatialLLM: From Multi-modality Data to Urban Spatial Intelligence](https://mlsanity.com/paper/2505.12703/) (2025)
- [Spatio-Temporal LLM: Reasoning about Environments and Actions](https://mlsanity.com/paper/2507.05258/) (2025)
- [TR-LLM: Integrating Trajectory Data for Scene-Aware LLM-Based Human Action Prediction](https://mlsanity.com/paper/2410.03993/) (2024)
- [Simulation-Assisted Optimization for Large-Scale Evacuation Planning with Congestion-Dependent Delays](https://mlsanity.com/paper/2209.01535/) (2022)
- [Navigating Motion Agents in Dynamic and Cluttered Environments through LLM Reasoning](https://mlsanity.com/paper/2503.07323/) (2025)
- [FABRIC: A Framework for the Design and Evaluation of Collaborative Robots with Extended Human Adaptation](https://mlsanity.com/paper/2104.01976/) (2021)

## Abstract (arXiv)

Anticipating human activities and their durations is essential in applications such as smart-home automation, simulation-based architectural and urban design, activity-based transportation system simulation, and human-robot collaboration, where adaptive systems must respond to human activities. Existing data-driven agent-based models--from rule-based to deep learning--struggle in low-data environments, limiting their practicality. This paper investigates whether large language models, pre-trained on broad human knowledge, can fill this gap by reasoning about everyday activities from compact contextual cues. We adopt a retrieval-augmented prompting strategy that integrates four sources of context--temporal, spatial, behavioral history, and persona--and evaluate it on the CASAS Aruba smart-home dataset. The evaluation spans two complementary tasks: next-activity prediction with duration estimation, and multi-step daily sequence generation, each tested with various numbers of few-shot examples provided in the prompt. Analyzing few-shot effects reveals how much contextual supervision is sufficient to balance data efficiency and predictive accuracy, particularly in low-data environments. Results show that large language models exhibit strong inherent temporal understanding of human behavior: even in zero-shot settings, they produce coherent daily activity predictions, while adding one or two demonstrations further refines duration calibration and categorical accuracy. Beyond a few examples, performance saturates, indicating diminishing returns. Sequence-level evaluation confirms consistent temporal alignment across few-shot conditions. These findings suggest that pre-trained language models can serve as promising temporal reasoners, capturing both recurring routines and context-dependent behavioral variations, thereby strengthening the behavioral modules of agent-based models.