# Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression

arXiv: https://arxiv.org/abs/2608.04569 · HTML: https://mlsanity.com/paper/2608.04569/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Prompt and context compression, one of 12 directions in Efficient LLM inference and serving.
- Builds on 32 of its 39 references that are in the map.
- First posted on arXiv 5 Aug 2026.

## Placed in

- Efficient LLM inference and serving / [Prompt and context compression](https://mlsanity.com/inference/prompt-and-context-compression/)

## Builds on (most cited first, 30 of 32)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps](https://mlsanity.com/paper/2011.01060/) (2020)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [MuSiQue: Multihop Questions via Single-hop Question Composition](https://mlsanity.com/paper/2108.00573/) (2021)
- [On Faithfulness and Factuality in Abstractive Summarization](https://mlsanity.com/paper/2005.00661/) (2020)
- [Large Language Models Can Be Easily Distracted by Irrelevant Context](https://mlsanity.com/paper/2302.00093/) (2023)
- [LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression](https://mlsanity.com/paper/2310.06839/) (2023)
- [Learning to Compress Prompts with Gist Tokens](https://mlsanity.com/paper/2304.08467/) (2023)
- [Adapting Language Models to Compress Contexts](https://mlsanity.com/paper/2305.14788/) (2023)
- [LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression](https://mlsanity.com/paper/2403.12968/) (2024)
- [End-to-end Neural Coreference Resolution](https://mlsanity.com/paper/1707.07045/) (2017)
- [Long Context Compression with Activation Beacon](https://mlsanity.com/paper/2401.03462/) (2024)
- [Prompt Compression with Context-Aware Sentence Encoding for Fast and Improved LLM Inference](https://mlsanity.com/paper/2409.01227/) (2024)
- [Extending Context Window of Large Language Models via Semantic Compression](https://mlsanity.com/paper/2312.09571/) (2023)
- [Discrete Prompt Compression with Reinforcement Learning](https://mlsanity.com/paper/2308.08758/) (2023)
- [An efficient encoder-decoder architecture with top-down attention for speech separation](https://mlsanity.com/paper/2209.15200/) (2022)
- [Learning-Based Single-Document Summarization with Compression and Anaphoricity Constraints](https://mlsanity.com/paper/1603.08887/) (2016)
- [TCRA-LLM: Token Compression Retrieval Augmented Large Language Model for Inference Cost Reduction](https://mlsanity.com/paper/2310.15556/) (2023)
- [Perception Compressor: A Training-Free Prompt Compression Framework in Long Context Scenarios](https://mlsanity.com/paper/2409.19272/) (2024)
- [TACO-RL: Task Aware Prompt Compression Optimization with Reinforcement Learning](https://mlsanity.com/paper/2409.13035/) (2024)
- [Fundamental Limits of Prompt Compression: A Rate-Distortion Framework for Black-Box Language Models](https://mlsanity.com/paper/2407.15504/) (2024)
- [Characterizing Prompt Compression Methods for Long Context Inference](https://mlsanity.com/paper/2407.08892/) (2024)
- [A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression](https://mlsanity.com/paper/2412.17483/) (2024)
- [Understanding and Improving Information Preservation in Prompt Compression for LLMs](https://mlsanity.com/paper/2503.19114/) (2025)
- [Subnetwork-to-go: Elastic Neural Network with Dynamic Training and Customizable Inference](https://mlsanity.com/paper/2312.03464/) (2023)
- [From Reading to Compressing: Exploring the Multi-document Reader for Prompt Compression](https://mlsanity.com/paper/2410.04139/) (2024)
- [DAC: A Dynamic Attention-aware Approach for Task-Agnostic Prompt Compression](https://mlsanity.com/paper/2507.11942/) (2025)

## Abstract (arXiv)

Hard prompt compression reduces long-context inference cost by independently scoring tokens, sentences, or chunks and retaining the highest-scoring units under a budget. We identify a structural failure in this procedure: independent selection can split dependent evidence pairs, retaining one member while deleting the other. When retained text contains an answer but deleted text defines the entity needed to interpret it, we call the result referential dangling. At a compression ratio of 0.30, Beaver, which ranks coherent chunks using Qwen3-0.6B embeddings, leaves the answer path incomplete in 34-54% of bridge examples across three multi-hop question answering datasets. On a shared HotpotQA bridge set, all six hard compressors we test exhibit dangling at rates up to 60%, and every document in LongBench-v2 Single-Document QA contains at least one dangling reference. On dangling examples evaluated with Qwen3-8B, reinserting the missing supporting paragraph while removing nonsupporting paragraphs to maintain the token budget improves accuracy by 29-34 percentage points (p < 0.0001), recovering at least 88% of the gap to contexts retaining both supporting paragraphs. Stronger answer models do not absorb the loss: on MuSiQue, GPT-5.5 is 8.8 points less accurate on compressed contexts than on contexts retaining both supporting paragraphs. Finally, we train a compact classifier to rank omitted sentences by whether they are needed to interpret retained text and reinsert the top-ranked candidates without support annotations at inference. On HotpotQA with Qwen3-8B, this automatic restoration improves accuracy by 4.7 points while changing the compression ratio only from 0.30 to 0.31. Hard compressors should optimize both relevance and referential completeness.