Multi-Turn Conditional Instruction State Tracking Dataset
#condition state tracking #instruction update reasoning #rule invalidation detection #multi-turn dialogue management #conditional instruction following #conversational state tracking
- 3840
- 0.5GB
- JSON
- CATL
MOBIUSI INC
Updated:2026-10-25
AI Analysis & Value Prop
This dataset contains multi-turn text conversations in which users establish conditional rules and later add to them, revise them, or invalidate them. Each sample records the full conversation and annotations of the effective rules and expected assistant response at each turn, supporting training in persistent instruction tracking and rule-aware response generation. It is suitable for conditional instruction-following training, dialogue generation, and evaluation of multi-turn instruction handling.
Dataset Insights
Sample Examples
{}
Technical Specifications
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| turn_annotations | array | Annotations for assistant response turns that describe the effective rules and expected response at each point in the conversation. |
| conversation_text | string | The complete source conversation, including user instructions, later rule updates or invalidations, and assistant messages. |
Compliance Statement
| Authorization Type | Proprietary - Commercial AI Training License (No Redistribution) |
| Commercial Use | Requires exclusive subscription or authorization contract (monthly or per-invocation charging) |
| Privacy and Anonymization | No PII, no real company names, simulated scenarios follow industry standards |
| Compliance System | Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs |
Frequently Asked Questions
- What does the dataset contain?
- Samples include complete multi-turn conversations covering user-defined conditional rules, later additions or revisions, and assistant responses. Turn annotations describe the active rules and expected response at each stage.
- Which capabilities can it support for training or evaluation?
- It can train models to track conditional state across a conversation, determine whether rules remain active, and generate text that follows the current rules. It can also support evaluation of multi-turn conditional instruction following.
- What are the dataset size and file specifications?
- The available information does not specify the number of samples, file format, or partitioning. Refer to the latest product details or delivery documentation for these specifications.
- How can I obtain the data, and what are the licensing terms?
- Check the dataset product page or contact the provider for access procedures, permitted uses, and commercial licensing terms. Review the applicable license and delivery documentation before use.
Can't find the data you need?
Post a request and let data providers reach out to you.
Get this Dataset
Verified for Enterprise UseCite this Work
@dataset{Mobiusi2026,
title={Multi-Turn Conditional Instruction State Tracking Dataset},
author={MOBIUSI INC},
year={2026},
url={https://www.mobiusi.com/datasets/0dc002d7bf227537bfcbe6e6cd59aac6},
urldate={2026-10-22},
keywords={multi-turn instruction following,conditional instruction dataset,instruction state tracking,multi-turn dialogue data,text generation training data,rule update conversations,LLM training data,instruction following evaluation},
version={1.0}
}Using this in research? Please cite us.
placeholder
placeholder
placeholder
placeholder
placeholder
placeholder
placeholder