# DeCAL Tokenwise Compression

arXiv: https://arxiv.org/abs/2508.08514 · HTML: https://mlsanity.com/paper/2508.08514/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 29 of its 32 references that are in the map.
- First posted on arXiv 11 Aug 2025.

## Builds on (most cited first, 29 of 29)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [MS MARCO: A Human Generated MAchine Reading COmprehension Dataset](https://mlsanity.com/paper/1611.09268/) (2016)
- [BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models](https://mlsanity.com/paper/2104.08663/) (2021)
- [ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT](https://mlsanity.com/paper/2004.12832/) (2020)
- [Perceiver: General Perception with Iterative Attention](https://mlsanity.com/paper/2103.03206/) (2021)
- [Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond](https://mlsanity.com/paper/1602.06023/) (2016)
- [Efficient Transformers: A Survey](https://mlsanity.com/paper/2009.06732/) (2020)
- [Compressive Transformers for Long-Range Sequence Modelling](https://mlsanity.com/paper/1911.05507/) (2019)
- [A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents](https://mlsanity.com/paper/1804.05685/) (2018)
- [Adapting Language Models to Compress Contexts](https://mlsanity.com/paper/2305.14788/) (2023)
- [CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation](https://mlsanity.com/paper/2103.06874/) (2021)
- [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://mlsanity.com/paper/2112.07916/) (2021)
- [In-context Autoencoder for Context Compression in a Large Language Model](https://mlsanity.com/paper/2307.06945/) (2023)
- [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://mlsanity.com/paper/2006.03236/) (2020)
- [Byte Latent Transformer: Patches Scale Better Than Tokens](https://mlsanity.com/paper/2412.09871/) (2024)
- [MediaSum: A Large-scale Media Interview Dataset for Dialogue Summarization](https://mlsanity.com/paper/2103.06410/) (2021)
- [Hierarchical Transformers Are More Efficient Language Models](https://mlsanity.com/paper/2110.13711/) (2021)
- [PLAID: An Efficient Engine for Late Interaction Retrieval](https://mlsanity.com/paper/2205.09707/) (2022)
- [Scaling Transformer to 1M tokens and beyond with RMT](https://mlsanity.com/paper/2304.11062/) (2023)
- [A Survey on Transformer Compression](https://mlsanity.com/paper/2402.05964/) (2024)
- [Nugget: Neural Agglomerative Embeddings of Text](https://mlsanity.com/paper/2310.01732/) (2023)
- [Pre-computed memory or on-the-fly encoding? A hybrid approach to retrieval augmentation makes the most of your compute](https://mlsanity.com/paper/2301.10448/) (2023)
- [Multi-Vector Retrieval as Sparse Alignment](https://mlsanity.com/paper/2211.01267/) (2022)

## Abstract (arXiv)

This paper introduces DeCAL, a new method for tokenwise compression. DeCAL uses an encoder-decoder language model pretrained with denoising to learn to produce high-quality, general-purpose compressed representations from the encoder. DeCAL applies small modifications to the encoder, with the emphasis on maximizing compression quality, even at the expense of compute. We show that DeCAL at 2x compression can match uncompressed on several downstream tasks, with usually only a minor dropoff in metrics up to 8x compression, among question-answering, summarization, and multi-vector retrieval tasks. DeCAL offers significant savings where pre-computed dense representations can be utilized, and we believe the approach can be further developed to be more broadly applicable.