---
title: "Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A"
description: "Feeding large, changing bodies of knowledge to LLMs is hard. 34 references. Long prompts are slow and bounded by the context window."
authors: ["Wenxuan Xie", "Xuhong Wang"]
year: 2026
venue: "Findings of the Association for Computational Linguistics: ACL 2026"
source: "https://aclanthology.org/2026.findings-acl.1897/"
doi: "10.18653/v1/2026.findings-acl.1897"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/xie-2026-decoupled-reasoning-implicit-fact-tokens/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/xie-2026-decoupled-reasoning-implicit-fact-tokens/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/29/) / Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A Dual-Model Framework for Efficient Long-Context Inference

# Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A Dual-Model Framework for Efficient Long-Context Inference

**Authors:** [Wenxuan Xie](https://md.research.elamur.ai/free/en/authors/wenxuan-xie/) and [Xuhong Wang](https://md.research.elamur.ai/free/en/authors/xuhong-wang/) · **Year:** 2026 · **Venue:** [Findings of the Association for Computational Linguistics: ACL 2026](https://md.research.elamur.ai/free/en/venues/findings/) ([Findings of the Association for Computational Linguistics 2026](https://md.research.elamur.ai/free/en/venues/findings/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.findings-acl.1897/) · [PDF](https://aclanthology.org/2026.findings-acl.1897.pdf) · [DOI 10.18653/v1/2026.findings-acl.1897](https://doi.org/10.18653/v1/2026.findings-acl.1897) · [OpenAlex](https://openalex.org/W7166810593)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Feeding large, changing bodies of knowledge to LLMs is hard. Long prompts are slow and bounded by the context window. RAG depends on retriever quality. Editing weights risks catastrophic forgetting. Earlier soft-compression methods also tend to compress text without regard to the question, so evidence the task needs can vanish.

DRIFT addresses this by splitting the work between two models. A small knowledge model reads document chunks in parallel, together with the query, and condenses each chunk into a few latent "fact tokens." These are projected into the embedding space of a larger reasoning model, which answers using them instead of the raw text. Training uses a bucketed scheme: output size depends on input-length ranges rather than a fixed ratio. There are three objectives (LFRP, QAFT-DC, QAFT-QA) and a new Document–QA–Evidence dataset.

At strong compression, DRIFT beats similarly sized baselines on long-context benchmarks and cuts latency sharply. This makes long-document inference cheaper without retraining the reasoner on new facts.

## Key points
- With a Mistral 7B reasoner and 32× compression, LongBench v2 accuracy goes from 20.87% to 29.22%. Results stay competitive at 64× and 128×.
- End-to-end inference on 256k-token documents is about 7× faster on average than full-context processing, measured by TTFT.
- The authors release a Document–QA–Evidence training set of over 300K instances, with documents of 1K–8K tokens. Code is at github.com/Lancelot-Xie/DRIFT.
- In the ablation, QAFT-QA matters most. Removing it lowers BAMBOO from 60.15 to 45.14, LongBench v2 from 29.22 to 18.05, and LoCoMo from 57.73 to 36.89.
- Several size pairings work: 3B+7B, 1.5B+7B and 3B+14B. A larger reasoner helps modestly, while a smaller knowledge model changes little. Stated limitations are that tests stop at 14B parameters and that the latent fact tokens are hard to interpret.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Shanghai Artificial Intelligence Laboratory](https://md.research.elamur.ai/free/en/labs/shanghai-artificial-intelligence-laboratory/) — Research institute, Shanghai, China · 152 works · 13,166 citations

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 3,486 works · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,730 works

🧭 Topics (1): [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,685 works

## Cite

Cite as: Wenxuan Xie and Xuhong Wang. 2026. Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A Dual-Model Framework for Efficient Long-Context Inference. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{xie2026decoupled,
  title = {{Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A Dual-Model Framework for Efficient Long-Context Inference}},
  author = {Wenxuan Xie and Xuhong Wang},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1897},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1897}
}
```

- [← previous Deconstructing Instruction-Following: A New Benchmark for Granular Evaluation of Large Language Model Instruction Compliance Abilities](https://md.research.elamur.ai/free/en/papers/purpura-2026-deconstructing-instruction-following-new-benchmark/)
- [next → Decoupling Generalization and Adaptation in Meta-Learning for Large Language Models](https://md.research.elamur.ai/free/en/papers/vetcha-2026-decoupling-generalization-adaptation-meta-learning/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/29/)

## References

34 references, 19 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Chenxin An, Shansan Gong, Ming Zhong, Xingjian Zhao, Mukai Li, Jun Zhang, Lingpeng Kong, and Xipeng Qiu. 2024. [L-eval: Instituting standardized evaluation for long context language models](https://md.research.elamur.ai/free/en/papers/an-2023-l-eval-instituting-standardized-evaluation/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pape…
2. Yushi Bai, Shangqing Tu, Jiajie Zhang, Hao Peng, Xiaozhi Wang, Xin Lv, Shulin Cao, Jiazheng Xu, Lei Hou, Yuxiao Dong, Jie Tang, and Juanzi Li. 2025. [LongBench v2: Towards deeper understanding and reasoning on realistic long-context multitasks](https://md.research.elamur.ai/free/en/papers/bai-2024-longbench-v2-deeper-understanding/). In Proceedings of the 63rd Annual Meeting of the Associa…
3. Jiaqi Cao, Jiarui Wang, Rubin Wei, Qipeng Guo, Kai Chen, Bowen Zhou, and Zhouhan Lin. 2025. Memory decoder: A pretrained, plug-and-play memory for large language models. CoRR, abs/2508.09874. — [arXiv:2508.09874](https://arxiv.org/abs/2508.09874) [find](https://scholar.google.com/scholar?q=Jiaqi+Cao%2C+Jiarui+Wang%2C+Rubin+Wei%2C+Qipeng+Guo%2C+Kai+Chen%2C+Bowen+Zhou%2C+and+Zhouhan+Lin%2E+2025%2E+Memory+decoder%3A+A+pretrained%2C+plug%2Dand%2Dplay+memory+for+large+language+models%2E+CoRR%2C+abs%2F2508%2E09874%2E+%E2%80%94+%5BarXiv)
4. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde de Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, Alex Ray, Raul Puri, Gretchen Krueger, Michael Petrov, Heidy Khlaaf, Girish Sastry, Pamela Mishkin, Brooke Chan, Scott Gray, and 39 others. 202… — [Evaluating Large Language Models Trained on Code](https://md.research.elamur.ai/free/en/papers/chen-2021-evaluating-large-language-models-trained/)
5. Xin Cheng, Xun Wang, Xingxing Zhang, Tao Ge, Si-Qing Chen, Furu Wei, Huishuai Zhang, and Dongyan Zhao. 2024. xrag: Extreme context compression for retrieval-augmented generation with one token. In Advances in Neural Information Processing Systems, volume 37, pages 109487–109516. Curran Associates, I… [find](https://scholar.google.com/scholar?q=Xin+Cheng%2C+Xun+Wang%2C+Xingxing+Zhang%2C+Tao+Ge%2C+Si%2DQing+Chen%2C+Furu+Wei%2C+Huishuai+Zhang%2C+and+Dongyan+Zhao%2E+2024%2E+xrag%3A+Extreme+context+compression+for+retrieval%2Daugmented+generation+with+one+token%2E+In+Adv)
6. Xin Cheng, Wangding Zeng, Damai Dai, Qinyu Chen, Bingxuan Wang, Zhenda Xie, Kezhao Huang, Xingkai Yu, Zhewen Hao, Yukun Li, and 1 others. 2026. Conditional memory via scalable lookup: A new axis of sparsity for large language models. arXiv preprint arXiv:2601.07372. [find](https://scholar.google.com/scholar?q=Xin+Cheng%2C+Wangding+Zeng%2C+Damai+Dai%2C+Qinyu+Chen%2C+Bingxuan+Wang%2C+Zhenda+Xie%2C+Kezhao+Huang%2C+Xingkai+Yu%2C+Zhewen+Hao%2C+Yukun+Li%2C+and+1+others%2E+2026%2E+Conditional+memory+via+scalable+lookup%3A+A+new+axis+of+sp)
7. Alexis Chevalier, Alexander Wettig, Anirudh Ajith, and Danqi Chen. 2023. [Adapting language models to compress contexts](https://md.research.elamur.ai/free/en/papers/chevalier-2023-adapting-language-models-compress-contexts/). In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 3829–3846, Singapore. Association for Computational Linguistics.
8. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). CoRR, abs/2110.14168.
9. Zican Dong, Tianyi Tang, Junyi Li, Wayne Xin Zhao, and Ji-Rong Wen. 2024. [BAMBOO: A comprehensive benchmark for evaluating long text modeling capacities of large language models](https://md.research.elamur.ai/free/en/papers/dong-2024-bamboo-comprehensive-benchmark-evaluating-long/). In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation…
10. Tao Ge, Hu Jing, Lei Wang, Xun Wang, Si-Qing Chen, and Furu Wei. 2024. [In-context autoencoder for context compression in a large language model](https://md.research.elamur.ai/free/en/papers/ge-2023-context-autoencoder-context-compression-large/). In International Conference on Learning Representations, volume 2024, pages 2591–2607.
11. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. [LoRA: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). In International Conference on Learning Representations.
12. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix,… [find](https://scholar.google.com/scholar?q=Albert+Q%2E+Jiang%2C+Alexandre+Sablayrolles%2C+Arthur+Mensch%2C+Chris+Bamford%2C+Devendra+Singh+Chaplot%2C+Diego+de+las+Casas%2C+Florian+Bressand%2C+Gianna+Lengyel%2C+Guillaume+Lample%2C+Lucile+Saulnier%2C+L%C3%A9lio+Renard+Lav)
13. Yucheng Li, Bo Dong, Frank Guerin, and Chenghua Lin. 2023b. [Compressing context to enhance inference efficiency of large language models](https://md.research.elamur.ai/free/en/papers/li-2023-compressing-context-enhance-inference-efficiency/). In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6342–6353, Singapore. Association for Computational Linguistics.
14. Zihan Liao, Jun Wang, Hang Yu, Lingxiao Wei, Jianguo Li, Jun Wang, and Wei Zhang. 2025. E2LLM: Encoder elongated large language models for long-context understanding and reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 19201–19230, Suzhou, C… [find](https://scholar.google.com/scholar?q=Zihan+Liao%2C+Jun+Wang%2C+Hang+Yu%2C+Lingxiao+Wei%2C+Jianguo+Li%2C+Jun+Wang%2C+and+Wei+Zhang%2E+2025%2E+E2LLM%3A+Encoder+elongated+large+language+models+for+long%2Dcontext+understanding+and+reasoning%2E+In+Proceedings+of+t)
15. Fanfan Liu and Haibo Qiu. 2025. Context cascade compression: Exploring the upper limits of text compression. arXiv preprint arXiv:2511.15244. [find](https://scholar.google.com/scholar?q=Fanfan+Liu+and+Haibo+Qiu%2E+2025%2E+Context+cascade+compression%3A+Exploring+the+upper+limits+of+text+compression%2E+arXiv+preprint+arXiv%3A2511%2E15244%2E)
16. Huiqiang Jiang, Qianhui Wu, Xufang Luo, Dongsheng Li, Chin-Yew Lin, Yuqing Yang, and Lili Qiu. 2024. [LongLLMLingua: Accelerating and enhancing LLMs in long context scenarios via prompt compression](https://md.research.elamur.ai/free/en/papers/jiang-2023-longllmlingua-accelerating-enhancing-llms-long/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1:…
17. Adyasha Maharana, Dong-Ho Lee, Sergey Tulyakov, Mohit Bansal, Francesco Barbieri, and Yuwei Fang. 2024. Evaluating very long-term conversational memory of LLM agents. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 13851– 1387… [find](https://scholar.google.com/scholar?q=Adyasha+Maharana%2C+Dong%2DHo+Lee%2C+Sergey+Tulyakov%2C+Mohit+Bansal%2C+Francesco+Barbieri%2C+and+Yuwei+Fang%2E+2024%2E+Evaluating+very+long%2Dterm+conversational+memory+of+LLM+agents%2E+In+Proceedings+of+the+62nd+Annual)
18. Mingyu Jin, Weidi Luo, Sitao Cheng, Xinyi Wang, Wenyue Hua, Ruixiang Tang, William Yang Wang, and Yongfeng Zhang. 2025. [Disentangling memory and reasoning ability in large language models](https://md.research.elamur.ai/free/en/papers/jin-2025-disentangling-memory-reasoning-ability-large/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pape…
19. Jesse Mu, Xiang Li, and Noah Goodman. 2023. [Learning to compress prompts with gist tokens](https://md.research.elamur.ai/free/en/papers/mu-2023-learning-compress-prompts-gist-tokens/). In Advances in Neural Information Processing Systems, volume 36, pages 19327–19352. Curran Associates, Inc.
20. Yuri Kuratov, Aydar Bulatov, Petr Anokhin, Ivan Rodkin, Dmitry Igorevich Sorokin, Artyom Sorokin, and Mikhail Burtsev. 2024. [BABILong: Testing the limits of LLMs with long context reasoning-in-a-haystack](https://md.research.elamur.ai/free/en/papers/kuratov-2024-babilong-testing-limits-llms-long/). In The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks…
21. Yanlin Lai, Mitt Huang, Hangyu Guo, Xiangfeng Wang, Haodong Li, Shaoxiong Zhan, Liang Zhao, Chengyuan Yao, Yinmin Zhang, Qi Han, Chun Yuan, Zheng Ge, Xiangyu Zhang, and Daxin Jiang. 2026. R-align: Enhancing generative reward models through rationale-centric meta-judging. CoRR, abs/2602.06763. — [arXiv:2602.06763](https://arxiv.org/abs/2602.06763) [find](https://scholar.google.com/scholar?q=Yanlin+Lai%2C+Mitt+Huang%2C+Hangyu+Guo%2C+Xiangfeng+Wang%2C+Haodong+Li%2C+Shaoxiong+Zhan%2C+Liang+Zhao%2C+Chengyuan+Yao%2C+Yinmin+Zhang%2C+Qi+Han%2C+Chun+Yuan%2C+Zheng+Ge%2C+Xiangyu+Zhang%2C+and+Daxin+Jiang%2E+2026%2E+R%2Dalign%3A+Enh)
22. LangChain. 2025. langchain-text-splitters. PyPI package. Version 0.3.8; text splitting utilities. [find](https://scholar.google.com/scholar?q=LangChain%2E+2025%2E+langchain%2Dtext%2Dsplitters%2E+PyPI+package%2E+Version+0%2E3%2E8%3B+text+splitting+utilities%2E)
23. Junyi Li, Xiaoxue Cheng, Xin Zhao, Jian-Yun Nie, and Ji-Rong Wen. 2023a. [HaluEval: A large-scale hallucination evaluation benchmark for large language models](https://md.research.elamur.ai/free/en/papers/li-2023-halueval-large-scale-hallucination-evaluation/). In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6449–6464, Singapore. Association for Comput…
24. Zhuoshi Pan, Qianhui Wu, Huiqiang Jiang, Menglin Xia, Xufang Luo, Jue Zhang, Qingwei Lin, Victor Rühle, Yuqing Yang, Chin-Yew Lin, H. Vicky Zhao, Lili Qiu, and Dongmei Zhang. 2024. [LLMLingua-2: Data distillation for efficient and faithful task-agnostic prompt compression](https://md.research.elamur.ai/free/en/papers/pan-2024-llmlingua-2-data-distillation-efficient/). In Findings of the Associat…
25. Valentina Pyatkin, Saumya Malik, Victoria Graf, Hamish Ivison, Shengyi Huang, Pradeep Dasigi, Nathan Lambert, and Hannaneh Hajishirzi. 2025. [Generalizing verifiable instruction following](https://md.research.elamur.ai/free/en/papers/pyatkin-2025-generalizing-verifiable-instruction-following/). In The Thirty-ninth Annual Conference on Neural Information Processing Systems Datasets and Benchmarks Track.
26. Biqing Qi, Xinquan Chen, Junqi Gao, Dong Li, Jianxing Liu, Ligang Wu, and Bowen Zhou. 2024. Interactive continual learning: Fast and slow thinking. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12882–12892. [find](https://scholar.google.com/scholar?q=Biqing+Qi%2C+Xinquan+Chen%2C+Junqi+Gao%2C+Dong+Li%2C+Jianxing+Liu%2C+Ligang+Wu%2C+and+Bowen+Zhou%2E+2024%2E+Interactive+continual+learning%3A+Fast+and+slow+thinking%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Compute)
27. David Rau, Shuai Wang, Hervé Déjean, Stéphane Clinchant, and Jaap Kamps. 2025. Context embeddings for efficient answer generation in retrieval-augmented generation. In Proceedings of the Eighteenth ACM International Conference on Web Search and Data Mining, WSDM ’25, page 493–502, New York, NY, USA.… [find](https://scholar.google.com/scholar?q=David+Rau%2C+Shuai+Wang%2C+Herv%C3%A9+D%C3%A9jean%2C+St%C3%A9phane+Clinchant%2C+and+Jaap+Kamps%2E+2025%2E+Context+embeddings+for+efficient+answer+generation+in+retrieval%2Daugmented+generation%2E+In+Proceedings+of+the+Eighteenth+AC)
28. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R. Bowman. 2024. [GPQA: A graduate-level google-proof q&a benchmark](https://md.research.elamur.ai/free/en/papers/rein-2023-gpqa-graduate-level-google-proof/). In First Conference on Language Modeling.
29. Shivam Shandilya, Menglin Xia, Supriyo Ghosh, Huiqiang Jiang, Jue Zhang, Qianhui Wu, Victor Rühle, and Saravan Rajmohan. 2025. TACO-RL: Task aware prompt compression optimization with reinforcement learning. In Findings of the Association for Computational Linguistics: ACL 2025, pages 1582–1597, Vie… [find](https://scholar.google.com/scholar?q=Shivam+Shandilya%2C+Menglin+Xia%2C+Supriyo+Ghosh%2C+Huiqiang+Jiang%2C+Jue+Zhang%2C+Qianhui+Wu%2C+Victor+R%C3%BChle%2C+and+Saravan+Rajmohan%2E+2025%2E+TACO%2DRL%3A+Task+aware+prompt+compression+optimization+with+reinforcement+le)
30. Mirac Suzgun, Nathan Scales, Nathanael Schärli, Sebastian Gehrmann, Yi Tay, Hyung Won Chung, Aakanksha Chowdhery, Quoc Le, Ed Chi, Denny Zhou, and Jason Wei. 2023. [Challenging BIG-bench tasks and whether chain-of-thought can solve them](https://md.research.elamur.ai/free/en/papers/suzgun-2023-challenging-big-bench-tasks-whether/). In Findings of the Association for Computational Linguistics: A…
31. Jason Wei, Maarten Bosma, Vincent Zhao, Kelvin Guu, Adams Wei Yu, Brian Lester, Nan Du, Andrew M. Dai, and Quoc V Le. 2022. [Finetuned language models are zero-shot learners](https://md.research.elamur.ai/free/en/papers/wei-2021-finetuned-language-models-zero-shot/). In International Conference on Learning Representations.
32. Rubin Wei, Jiaqi Cao, Jiarui Wang, Jushi Kai, Qipeng Guo, Bowen Zhou, and Zhouhan Lin. 2025. MLP memory: A retriever-pretrained memory for large language models. CoRR, abs/2508.01832. — [arXiv:2508.01832](https://arxiv.org/abs/2508.01832) [find](https://scholar.google.com/scholar?q=Rubin+Wei%2C+Jiaqi+Cao%2C+Jiarui+Wang%2C+Jushi+Kai%2C+Qipeng+Guo%2C+Bowen+Zhou%2C+and+Zhouhan+Lin%2E+2025%2E+MLP+memory%3A+A+retriever%2Dpretrained+memory+for+large+language+models%2E+CoRR%2C+abs%2F2508%2E01832%2E+%E2%80%94+%5BarXiv%3A2508%2E01)
33. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, and 22 others. 2024. Qwen2.5 technical report. CoRR, abs/2412.1511… [find](https://scholar.google.com/scholar?q=An+Yang%2C+Baosong+Yang%2C+Beichen+Zhang%2C+Binyuan+Hui%2C+Bo+Zheng%2C+Bowen+Yu%2C+Chengyuan+Li%2C+Dayiheng+Liu%2C+Fei+Huang%2C+Haoran+Wei%2C+Huan+Lin%2C+Jian+Yang%2C+Jianhong+Tu%2C+Jianwei+Zhang%2C+Jianxin+Yang%2C+Jiaxi+Yang%2C+Jin)
34. Jingwei Yi, Yueqi Xie, Bin Zhu, Emre Kiciman, Guangzhong Sun, Xing Xie, and Fangzhao Wu. 2025. Benchmarking and defending against indirect prompt injection attacks on large language models. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1, KDD ’25, page 180… [find](https://scholar.google.com/scholar?q=Jingwei+Yi%2C+Yueqi+Xie%2C+Bin+Zhu%2C+Emre+Kiciman%2C+Guangzhong+Sun%2C+Xing+Xie%2C+and+Fangzhao+Wu%2E+2025%2E+Benchmarking+and+defending+against+indirect+prompt+injection+attacks+on+large+language+models%2E+In+Proceed)

*Updated 2 November 2026*