Learning a Single Token to Replace Long System Prompts in LLMs

[2511.23271] Learning a Single Token to Replace Long System Prompts in LLMs Skip to main content Search Submit Donate Log in Computer Science > Computation and Language arXiv:2511.23271 (cs) [Submitted on 28 Nov 2025 ( v1 ), last revised 28 Aug 2026 (this version, v2)] Title: Learning a Single Token to Replace Long System Prompts in LLMs Authors: Jiancheng Dong , Pengyue Jia , Jingyu Peng , Maolin Wang , Yuhao Wang , Lixin Su , Xin Sun , Shuaiqiang Wang , Dawei Yin , Xiangyu Zhao View PDF HTML (experimental) Abstract: Long system prompts are widely used to steer Large Language Models (LLMs), but repeatedly processing them at inference time is inefficient and consumes valuable context budget. This motivates a central question: can the behavioral effect of a long system prompt be retained using only a minimal learned representation? To enable this, we propose a lightweight training framework that learns a single Behavior-Equivalent Token ([BE]). The framework first trains [BE] to encode the semantic content of the original system prompt via reconstruction, and then distills the prompt's downstream behavior into this single token. Importantly, our method requires no update to the pretrained LLM weights, no auxiliary compression models, and no labeled responses. Empirical evaluations on three datasets show that replacing long prompts with a single [BE] token yields up to a $3000\times$ prompt compression ratio, while retaining about 98% of the downstream performance of the original system prompts. This substantially reduces inference cost and frees nearly the entire context window for user inputs and model outputs. Comments: 12 pages, 4 figures Subjects: Computation and Language (cs.CL) Cite as: arXiv:2511.23271 [cs.CL] (or arXiv:2511.23271v2 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2511.23271 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Jiancheng Dong [ view email ] [v1] Fri, 28 Nov 2025 15:22:52 UTC (6,625 KB) [v2] Fri, 28 Aug 2026 06:59:02 UTC (4,612 KB) Full-text links: Access Paper: View PDF HTML (experimental) TeX Source view license Additional Features Audio Summary Current browse context: cs.CL < prev | next > new | recent | 2025-11 Change to browse by: cs References & Citations SciX Google Scholar Semantic Scholar export BibTeX citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab)

Related documents

• Learning a Single Token to Replace Long System Prompts in LLMs#682528
• Dynamic Deep Prompt Optimization for Defending Against Jailbreak Attacks on LLMs#1062086
• Configuration, Not Conscience: A Large-Scale Empirical Study of LLM System Prompts#1106274
• Teaching LLMs Program Semantics via Symbolic Execution Traces#168378
• Grammar-Aligned Decoding#925022