File size: 1,980 Bytes
047e215
 
 
 
 
 
01d914f
047e215
 
 
 
27b039d
 
83a3e9d
e57e016
27b039d
 
a57e0c3
 
 
 
 
881a313
e57e016
 
 
 
83a3e9d
 
 
 
 
 
 
 
 
e57e016
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
---
license: cc-by-4.0
pretty_name: MemoryTax-Res
language:
- en
tags:
- agent-memory
- llm-agents
- rag
---

# MemoryTax-Res

[![arXiv](https://img.shields.io/badge/arXiv-Paper-b31b1b?style=for-the-badge&logo=arxiv&logoColor=white)](https://arxiv.org/abs/2610.34073)  [![Project Page](https://img.shields.io/badge/Project-Page-1f8acb?style=for-the-badge&logo=googlechrome&logoColor=white)](https://hexuan21.github.io/MemoryTax/)  [![Code](https://img.shields.io/badge/Code-GitHub-181717?style=for-the-badge&logo=github&logoColor=white)](https://github.com/hexuan21/MemoryTax)  [![HF | Res Root Repo](https://img.shields.io/badge/HF-Res%20Root%20Repo-ffd21e?style=for-the-badge&logo=huggingface&logoColor=white&labelColor=000000)](https://huggingface.co/datasets/hexuan21/MemoryTax-Res)

Index of the MemoryTax result repos (this repo holds no data).

- [MemoryTax-Res-Qwen3-6-27B](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-Qwen3-6-27B) — 42.2 GB
- [MemoryTax-Res-Qwen3-5-4B](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-Qwen3-5-4B) — 49.1 GB
- [MemoryTax-Res-GPT-5-6-luna](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-GPT-5-6-luna) — 16.2 GB
- [MemoryTax-Res-experiential-mem](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-experiential-mem) — 0.1 GB
- [MemoryTax-Res-world-model-mem](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-world-model-mem) — 1.0 GB
- [MemoryTax-Res-Other](https://huggingface.co/datasets/hexuan21/MemoryTax-Res-Other) — 28.5 GB

## Citation

```bibtex
@misc{he2026memorytaxdissectingnet,
      title={Memory Tax: Dissecting the Net Value of Agent Memory},
      author={Xuan He and Sining Zhoubian and Ruozhen Yang and Yuhao Wen and Zhineng Wang and Yuyang Hu and Ke Yang and Yuyu Zhang and Ping Nie and Jiawei Han and ChengXiang Zhai},
      year={2026},
      eprint={2610.34073},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2610.34073},
}
```