Buckets:

Chetro983
/

memoryarena-bucket

Files

xet

Chetro983/memoryarena-bucket / README.md

Chetro983

4 months ago

preview code

download

raw

4.76 kB

---
configs:
  - config_name: bundled_shopping
    data_files:
      - split: test
        path: bundled_shopping/*.jsonl
  - config_name: progressive_search
    data_files:
      - split: test
        path: progressive_search/*.jsonl
  - config_name: group_travel_planner
    data_files:
      - split: test
        path: group_travel_planner/*.jsonl
  - config_name: formal_reasoning_math
    data_files:
      - split: test
        path: formal_reasoning_math/*.jsonl
  - config_name: formal_reasoning_phys
    data_files:
      - split: test
        path: formal_reasoning_phys/*.jsonl
---
# MemoryArena Dataset

## Overview
This dataset contains structured multi-session agentic tasks with question ```[list]```, answer ```[list]``` with necessary background context. Each row in the ```jsonl``` represents a agentic task ```[dict]``` with multiple subtasks, their corresponding answers, and background information.

## Dataset Structure

Each line in the JSONL file is a dictionary with the following fields:

- **id** (int): Unique identifier for each agentic task entry
- **questions** (list of str): List of sub-task queries
- **answers** (list of str): List of corresponding answers for each subtask
- **backgrounds** (str, or list of str): List of background/context information for each task
  - In Bundled Shopping and Progressive Search: no necessary backgrounds.
  - In Travel Planner: the travel details of the base person in each task serve as the background information for all subtasks.
  - In Formal Reasoning (Math and Phys): each subtask may has its background information. 

## Usage

Load with Hugging Face Datasets
```python
from datasets import load_dataset
ds = load_dataset("ZexueHe/memoryarena", "bundled_shopping")
ds = load_dataset("ZexueHe/memoryarena", "progressive_search")
ds = load_dataset("ZexueHe/memoryarena", "group_travel_planner")
ds = load_dataset("ZexueHe/memoryarena", "formal_reasoning_math")
ds = load_dataset("ZexueHe/memoryarena", "formal_reasoning_phys")
```

### Example task
In Bundled Webshop:
```json
{
  "id": 0,
  "questions": [
    "search subtask 1",
    "search subtask 1",
    ...
  ],
  "answers": [
    "search subtask result 1",
    "search subtask result 2",
    ...
  ]
}
```
In Progressive Search:
```json
{
  "id": 0,
  "questions": [
    "buy subtask item1",
    "buy subtask item 2",
    ...
  ],
  "answers": [
    {"target_asin": "B00TUDFEW2", "attributes": ["Almond Flour",...]},
    {"target_asin": "B08957C9ZH", "attributes": [...]},
    ...
  ]
}
```

In Group Travel Planner:

```json
{
  "id": 0,
  "base_person":{
    "name": "Jennifer",
    "query": "I am Jennifer. Please help me plan a trip from St. Petersburg to Rockford spanning 3 days...", # The travel requirements of the base person.
    "daily_plans": [ 
      {"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
      {"days": 2, "current_city": "Rockford", "transportation": "..."},
      ...
    ], # Detailed plans of the base person.
  },
  "questions": [
    "I am Eric.\n I'm joining Jennifer for this trip...",
    "I am Emma.\n I'm traveling with Jennifer and Eric... ",
    ...
  ], # subtask intructions
  "answers": [
   [ 
      {"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
      {"days": 2, "current_city": "Rockford", "transportation": "..."},
      ...
    ], # The answer of Subtask 1
    [ 
      {"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
      {"days": 2, "current_city": "Rockford", "transportation": "..."},
      ...
    ], # The answer of Subtask 2
    ....
  ]
}
```

In Formal Reasoning (Math and Phys)
```json
{
  "id": 0,
  "paper_name": "paper_id", # which paper the questions are created from 
  "backgrounds": [
    "necessary definitions, formulations, and relevant context of subtask 1",
    "necessary definitions, formulations, and relevant context of subtask 2",
    ...
  ]
  "questions": [
    "Math subtask question 1",
    "Math subtask question 2",
    ...
  ],
  "answers": [
    "Math result for subtask 1",
     "Math result for subtask 2",
    ...
  ]
}
```



## License
This dataset is licensed under the [Creative Commons Attribution 4.0 International (CC-BY-4.0)](https://creativecommons.org/licenses/by/4.0/) license.

## Citation
If you use this dataset, please cite:
```bibtex
@article{he2026memoryarena,
  title={MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks},
  author={He, Zexue and Wang, Yu and Zhi, Churan and Hu, Yuanzhe and Chen, Tzu-Ping and Yin, Lang and Chen, Ze and Wu, Tong Arthur and Ouyang, Siru and Wang, Zihan and others},
  journal={arXiv preprint arXiv:2602.16313},
  year={2026}
}
```

Xet Storage Details

Size:: 4.76 kB
Xet hash:: 2318d3adfe9e7037213609f69b26a7f519a5b1998a7ca4da1aedca64e61a2989

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.

	---
	configs:
	- config_name: bundled_shopping
	data_files:
	- split: test
	path: bundled_shopping/*.jsonl
	- config_name: progressive_search
	data_files:
	- split: test
	path: progressive_search/*.jsonl
	- config_name: group_travel_planner
	data_files:
	- split: test
	path: group_travel_planner/*.jsonl
	- config_name: formal_reasoning_math
	data_files:
	- split: test
	path: formal_reasoning_math/*.jsonl
	- config_name: formal_reasoning_phys
	data_files:
	- split: test
	path: formal_reasoning_phys/*.jsonl
	---

	# MemoryArena Dataset

	## Overview
	This dataset contains structured multi-session agentic tasks with question ```[list]```, answer ```[list]``` with necessary background context. Each row in the ```jsonl``` represents a agentic task ```[dict]``` with multiple subtasks, their corresponding answers, and background information.

	## Dataset Structure

	Each line in the JSONL file is a dictionary with the following fields:

	- id (int): Unique identifier for each agentic task entry
	- questions (list of str): List of sub-task queries
	- answers (list of str): List of corresponding answers for each subtask
	- backgrounds (str, or list of str): List of background/context information for each task
	- In Bundled Shopping and Progressive Search: no necessary backgrounds.
	- In Travel Planner: the travel details of the base person in each task serve as the background information for all subtasks.
	- In Formal Reasoning (Math and Phys): each subtask may has its background information.

	## Usage

	Load with Hugging Face Datasets
	```python
	from datasets import load_dataset

	ds = load_dataset("ZexueHe/memoryarena", "bundled_shopping")
	ds = load_dataset("ZexueHe/memoryarena", "progressive_search")
	ds = load_dataset("ZexueHe/memoryarena", "group_travel_planner")
	ds = load_dataset("ZexueHe/memoryarena", "formal_reasoning_math")
	ds = load_dataset("ZexueHe/memoryarena", "formal_reasoning_phys")
	```

	### Example task
	In Bundled Webshop:
	```json
	{
	"id": 0,
	"questions": [
	"search subtask 1",
	"search subtask 1",
	...
	],
	"answers": [
	"search subtask result 1",
	"search subtask result 2",
	...
	]
	}
	```
	In Progressive Search:
	```json
	{
	"id": 0,
	"questions": [
	"buy subtask item1",
	"buy subtask item 2",
	...
	],
	"answers": [
	{"target_asin": "B00TUDFEW2", "attributes": ["Almond Flour",...]},
	{"target_asin": "B08957C9ZH", "attributes": [...]},
	...
	]
	}
	```

	In Group Travel Planner:

	```json
	{
	"id": 0,
	"base_person":{
	"name": "Jennifer",
	"query": "I am Jennifer. Please help me plan a trip from St. Petersburg to Rockford spanning 3 days...", # The travel requirements of the base person.
	"daily_plans": [
	{"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
	{"days": 2, "current_city": "Rockford", "transportation": "..."},
	...
	], # Detailed plans of the base person.
	},
	"questions": [
	"I am Eric.\n I'm joining Jennifer for this trip...",
	"I am Emma.\n I'm traveling with Jennifer and Eric... ",
	...
	], # subtask intructions
	"answers": [
	[
	{"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
	{"days": 2, "current_city": "Rockford", "transportation": "..."},
	...
	], # The answer of Subtask 1
	[
	{"days": 1, "current_city": "from St. Petersburg to Rockford", "transportation": "..."},
	{"days": 2, "current_city": "Rockford", "transportation": "..."},
	...
	], # The answer of Subtask 2
	....
	]
	}
	```

	In Formal Reasoning (Math and Phys)
	```json
	{
	"id": 0,
	"paper_name": "paper_id", # which paper the questions are created from
	"backgrounds": [
	"necessary definitions, formulations, and relevant context of subtask 1",
	"necessary definitions, formulations, and relevant context of subtask 2",
	...
	]
	"questions": [
	"Math subtask question 1",
	"Math subtask question 2",
	...
	],
	"answers": [
	"Math result for subtask 1",
	"Math result for subtask 2",
	...
	]
	}
	```



	## License
	This dataset is licensed under the [Creative Commons Attribution 4.0 International (CC-BY-4.0)](https://creativecommons.org/licenses/by/4.0/) license.

	## Citation
	If you use this dataset, please cite:
	```bibtex
	@article{he2026memoryarena,
	title={MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks},
	author={He, Zexue and Wang, Yu and Zhi, Churan and Hu, Yuanzhe and Chen, Tzu-Ping and Yin, Lang and Chen, Ze and Wu, Tong Arthur and Ouyang, Siru and Wang, Zihan and others},
	journal={arXiv preprint arXiv:2602.16313},
	year={2026}
	}
	```