Datasets:

llm-semantic-router
/

feedback-detector-dataset

text stringlengths 7 100	label_name stringclasses 4 values
What would you like to know about dapsone?	NEED_CLARIFICATION
What would you like to know about da2pp?	NEED_CLARIFICATION
Are there any other buses available? leaving from Beach, CA.	WANT_DIFFERENT
Find me a salon	WANT_DIFFERENT
so, no! i already told you.	WRONG_ANSWER
I need more details about the code	NEED_CLARIFICATION
Can you find something else, preferably with United Airlines?	WANT_DIFFERENT
Nope thanks, that is not required.	WANT_DIFFERENT
Do you want to add another? [USER No, I don't need any more.	WANT_DIFFERENT
Thanks a lot! I'll try that.	SAT
What else? Maybe a concert on March 5th	WANT_DIFFERENT
What do you want to do with open windows?	NEED_CLARIFICATION
that was all for me. Thanks for your assistance.	WRONG_ANSWER
No, we want the Premium Economy class.	WRONG_ANSWER
No, you've taken care of the only item on my agenda for now.	WRONG_ANSWER
What would you like to know about macau?	NEED_CLARIFICATION
What would you like to know about frilled shark?	NEED_CLARIFICATION
No, I will ask it later. Now, I want to check the weather condition there on the 7th.	WANT_DIFFERENT
Can you find some other hotel I might like? I am interested for a 3 star hotel.	WANT_DIFFERENT
Can we use a differnet approach?	WANT_DIFFERENT
No for your time.	WRONG_ANSWER
What would you like to know about brooks automation?	NEED_CLARIFICATION
What do you want to do with optical drive?	NEED_CLARIFICATION
Tell me about the different options, look for 4.4 rating or even higher	WANT_DIFFERENT
No. Thanks alot need your help, though.	WRONG_ANSWER
No. That will be all. QUERY] you!	WRONG_ANSWER
No, cast it on my bedroom speaker QUERY]	WRONG_ANSWER
Okay, thanks for clarifying! 👍	SAT
What would you like to know about minsk?	NEED_CLARIFICATION
What exactly is the recommendation?	NEED_CLARIFICATION
Perfect, exactly what I needed!	SAT
What would you like to know about lexus nx?	NEED_CLARIFICATION
This part about this is unclear	NEED_CLARIFICATION
Yeah, we already that Is there anything directed by Hari Sama?	WANT_DIFFERENT
Got it, thanks! :)	SAT
No, that was helpful for me. Thanks!	WRONG_ANSWER
No. Thanks. That'll be all for now.	WRONG_ANSWER
What would you like to know about botswana?	NEED_CLARIFICATION
would you recommend other options?	WANT_DIFFERENT
Alright, That makes sense, thanks!	SAT
that's cool. are there any others?	WANT_DIFFERENT
What do you want to do with text messaging?	NEED_CLARIFICATION
Negative thanks, could you get me a house for 1 with a 4.5 rating or higher?	WANT_DIFFERENT
What would you like to know about fexofenadine?	NEED_CLARIFICATION
No thanks. That's all I need else now.	WRONG_ANSWER
Great, thank you! I'll try that.	SAT
What about Economy tickets for 2 people?	WANT_DIFFERENT
I actually need to buy tickets but maybe some other time.	WANT_DIFFERENT
Thanks for the help! Much appreciated.	SAT
What would you like to know about oasis petroleum?	NEED_CLARIFICATION
What do you want to do with a traffic ticket?	NEED_CLARIFICATION
no that's it, QUERY] you.	WRONG_ANSWER
Wonderful, thank you so much!	SAT
That works perfectly, thank you!	SAT
What do you want to do with laminate wood flooring?	NEED_CLARIFICATION
Leave it and search for some other cars that suits me.	WANT_DIFFERENT
That's perfect! :)	SAT
Great to know. Any other museum I could go to?	WANT_DIFFERENT
Ok, Great, this is useful!	SAT
Are there buses available from four people?	WANT_DIFFERENT
Yes, this is helpful!	SAT
What would you like to know about adinkra symbols?	NEED_CLARIFICATION
Nope, but your help was really appreciated.	WRONG_ANSWER
No, i need 2 tickets	WRONG_ANSWER
No, I want morning, not afternoon.	WRONG_ANSWER
That one is neat as well, but I'd rather you find me an event that starts on March 8th.	WANT_DIFFERENT
What do you want to do with washer dryer?	NEED_CLARIFICATION
I like that the bus is Economy but are there other buses available? I would like to go to Vegas.	WANT_DIFFERENT
there anything else? [USER No. Thanks. all.	WRONG_ANSWER
No, I need something else for two people.	WANT_DIFFERENT
Alright, Awesome, this is perfect!	SAT
What would you like to know about battle of midway?	NEED_CLARIFICATION
What do you want to know about the bird?	NEED_CLARIFICATION
No, the table should be for one person at 7 in the evening.	WRONG_ANSWER
Do you have any another options with less rent?	WANT_DIFFERENT
Yes, that's what I was looking for! 👍	SAT
Is there some other bus which you can recommend? I will leave for my trip on Sunday this week.	WANT_DIFFERENT
That works perfectly, thank you!	SAT
What do you want to know about quality health care?	NEED_CLARIFICATION
What do you want to do with calcium?	NEED_CLARIFICATION
What about the 9th? Any times available?	WANT_DIFFERENT
What do you want to know about spear physical therapy?	NEED_CLARIFICATION
anything else can help you with? [USER No, that is Thank	WRONG_ANSWER
Would you also check my savings balance?	WANT_DIFFERENT
What would you like to know about university of south florida?	NEED_CLARIFICATION
What do you want to do with jury duty?	NEED_CLARIFICATION
Okay, thanks for clarifying! I'll try that.	SAT
Awesome, appreciate the help! 👍	SAT
No actually, cast it in the my play	WRONG_ANSWER
Alright, Perfect, I understand now!	SAT
What would you like to know about koch industries?	NEED_CLARIFICATION
Yes, that answers my question! This helps a lot.	SAT
Awesome, appreciate the help! 👍	SAT
Actually, the method is wrong	WRONG_ANSWER
No, I'm all set. Thanks!	WRONG_ANSWER
Yes, that's correct! I'll try that.	SAT
Thanks for the help! Much appreciated.	SAT
What would you like to know about waste management inc?	NEED_CLARIFICATION
Okay, thanks for clarifying! Much appreciated.	SAT
No,Change it to Would 4th at 11:00.	WRONG_ANSWER

End of preview. Expand in Data Studio

Feedback Detector Dataset

A large-scale multilingual dataset for 4-class user feedback classification, labeled using GPT-OSS-120B on AMD MI300X GPU.

Dataset Description

This dataset contains 51,694 examples of user feedback classified into 4 categories:

Label	Description	Count	%
SAT	User is satisfied	8,649	17%
NEED_CLARIFICATION	User needs more information	16,179	31%
WRONG_ANSWER	System gave incorrect response	19,919	39%
WANT_DIFFERENT	User wants something different	6,947	13%

Splits

Split	Examples
Train	46,524
Validation	5,170

Data Sources

The dataset combines multiple public dialogue and complaint datasets:

Source	Type	Examples	Languages
consumer_complaints_medium	Complaints	~6,000	English
customer_complaints	Complaints	~2,000	English
multiwoz	Task-Oriented Dialogue	~3,000	English
sgd	Schema-Guided Dialogue	~3,000	English
inscit	Information-Seeking	~2,500	English
mimics	Search Clarification	~2,000	English
hazumi	Affirmative/Negative	~1,500	Japanese
turkish_complaints	Complaints	~1,000	Turkish

Labeling Process

All examples were labeled using OpenAI GPT-OSS-120B served via vLLM on AMD MI300X GPU with:

Temperature: 0 (deterministic)
Structured JSON output
Retry logic with exponential backoff
Garbage detection and circuit breaker pattern
Parallel processing with rate limiting

Hardware

Component	Specification
GPU	AMD Instinct MI300X
VRAM	192 GB HBM3
Framework	vLLM with ROCm
Model	OpenAI GPT-OSS-120B

Usage

from datasets import load_dataset

dataset = load_dataset("llm-semantic-router/feedback-detector-dataset")

# Access splits
train = dataset["train"]
val = dataset["validation"]

# Example
print(train[0])
# {'text': '...', 'label': 0, 'label_name': 'SAT', 'source': 'multiwoz'}

Schema

Each example contains:

text (str): The user feedback text
label (int): Numeric label (0-3)
label_name (str): Label name (SAT, NEED_CLARIFICATION, WRONG_ANSWER, WANT_DIFFERENT)
source (str): Original dataset source

Label Mapping

label2id = {
    "SAT": 0,
    "NEED_CLARIFICATION": 1,
    "WRONG_ANSWER": 2,
    "WANT_DIFFERENT": 3
}

Intended Use

Fine-tuning feedback detection models (ModernBERT, mmBERT, etc.)
User satisfaction classification in conversational AI
Customer service automation
Dialogue system evaluation

Citation

If you use this dataset, please cite:

@dataset{feedback_detector_dataset,
  title={Feedback Detector Dataset},
  author={LLM Semantic Router Team},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/llm-semantic-router/feedback-detector-dataset}
}

License

Apache 2.0

Downloads last month: 54

Size of downloaded dataset files:

421 kB

Size of the auto-converted Parquet files:

421 kB

Number of rows:

20,881

Models trained or fine-tuned on llm-semantic-router/feedback-detector-dataset