Activity counts measure attention, not model quality. Missing counts are unknown. Publication status follows the linked source; a listed venue alone does not confirm peer review.
- Form factor
- Other
- Signals
- IMU, Text
- License
none declared (no LICENSE file in repo at the recorded commit; all rights reserved by default)
- Type
- Task model
- Form factor
- Other
- Parameters
- Not reported
- Architecture
- Not reported
- Venue
- Not reported
- Year
- Not reported
- arXiv
- Not reported
- DOI
- Not reported
- Weights
- Not released
- Licence
- none declared (no LICENSE file in repo at the recorded commit; all rights reserved by default)
- Caveats
- Not reported
- Datasets
- Devices
- Added
- Oct 19, 2026
- Updated
- Oct 19, 2026
Original description
ZARA (Z̲ero-training A̲ctivity R̲easoning A̲gents) is a training-free, evidence-grounded LLM agent framework for motion time-series reasoning. It combines statistically grounded domain knowledge, multi-sensor retrieval-augmented evidence, and hierarchical multi-agent reasoning to classify motion sensor data with transparent, human-readable rationales—without fine-tuning or task-specific classifiers.
Before running the Gemini-based inference notebooks, set your API key as an environment variable:
Each dataset directory also includes preprocessing and feature-importance notebooks. Run preprocessing first, then feature-importance generation, and finally the Gemini inference notebook for the corresponding dataset.
Verbatim fromGitHub· Description from the repository README
Key results
How it works
Datasets linked in the hub; pretraining vs evaluation roles follow the paper.
- Inputs
- IMU, Text
- Approach
- code release · no weights
Devices· 1
Icons show the form factor: we have no openly licensed or manufacturer-cleared photo of these exact devices yet.
Leaderboard rows· 3
All leaderboards| Task | Dataset | Metric | Value | Rank |
|---|---|---|---|---|
| Human activity recognition (IMU) | MHEALTH | Macro-F1 ↑ | 86.1 (%) | #2 / 3 |
| Human activity recognition (IMU) | OPPORTUNITY | Macro-F1 ↑ | 92.5 (%) | #1 / 3 |
| Human activity recognition (IMU) | PAMAP2 | Macro-F1 ↑ | 76.9 (%) | #3 / 3 |
Datasets· 3
Papers· 1
Cite
APA
ZARA. (n.d.).
BibTeX
@misc{zarazara,
title = {{ZARA}}
}Built from the hub record. Check it against the publisher before submitting.
History
Added to the ChatPPG research hub