# CacheFormer: High Attention-Based Segment Caching

arXiv: https://arxiv.org/abs/2504.13981 · HTML: https://mlsanity.com/paper/2504.13981/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 30 references that are in the map.
- First posted on arXiv 18 Apr 2025.

## Builds on (most cited first, 25 of 25)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Efficiently Modeling Long Sequences with Structured State Spaces](https://mlsanity.com/paper/2111.00396/) (2021)
- [Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention](https://mlsanity.com/paper/2006.16236/) (2020)
- [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://mlsanity.com/paper/1901.02860/) (2019)
- [Reformer: The Efficient Transformer](https://mlsanity.com/paper/2001.04451/) (2020)
- [Rethinking Attention with Performers](https://mlsanity.com/paper/2009.14794/) (2020)
- [Generating Long Sequences with Sparse Transformers](https://mlsanity.com/paper/1904.10509/) (2019)
- [Linformer: Self-Attention with Linear Complexity](https://mlsanity.com/paper/2006.04768/) (2020)
- [Hungry Hungry Hippos: Towards Language Modeling with State Space Models](https://mlsanity.com/paper/2212.14052/) (2022)
- [Efficient Content-Based Sparse Attention with Routing Transformers](https://mlsanity.com/paper/2003.05997/) (2020)
- [Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention](https://mlsanity.com/paper/2502.11089/) (2025)
- [xLSTM: Extended Long Short-Term Memory](https://mlsanity.com/paper/2405.04517/) (2024)
- [Mega: Moving Average Equipped Gated Attention](https://mlsanity.com/paper/2209.10655/) (2022)
- [Long-Short Transformer: Efficient Transformers for Language and Vision](https://mlsanity.com/paper/2107.02192/) (2021)
- [General-purpose, long-context autoregressive modeling with Perceiver AR](https://mlsanity.com/paper/2202.07765/) (2022)
- [The NLP Cookbook: Modern Recipes for Transformer based Deep Learning Architectures](https://mlsanity.com/paper/2104.10640/) (2021)
- [Perceiving Longer Sequences With Bi-Directional Cross-Attention Transformers](https://mlsanity.com/paper/2402.12138/) (2024)
- [LaMemo: Language Modeling with Look-Ahead Memory](https://mlsanity.com/paper/2204.07341/) (2022)
- [LongVQ: Long Sequence Modeling with Vector Quantization on Structured Memory](https://mlsanity.com/paper/2404.11163/) (2024)
- [Long Sequence Modeling with Attention Tensorization: From Sequence to Tensor Learning](https://mlsanity.com/paper/2410.20926/) (2024)

## Abstract (arXiv)

Efficiently handling long contexts in transformer-based language models with low perplexity is an active area of research. Numerous recent approaches like Linformer, Longformer, Performer, and Structured state space models (SSMs)., have not fully resolved this problem. All these models strive to reduce the quadratic time complexity of the attention mechanism while minimizing the loss in quality due to the effective compression of the long context. Inspired by the cache and virtual memory principle in computers, where in case of a cache miss, not only the needed data is retrieved from the memory, but the adjacent data is also obtained, we apply this concept to handling long contexts by dividing it into small segments. In our design, we retrieve the nearby segments in an uncompressed form when high segment-level attention occurs at the compressed level. Our en-hancements for handling long context include aggregating four attention mechanisms consisting of short sliding window attention, long compressed segmented attention, dynamically retrieving top k high attention uncompressed segments, and overlapping segments in long segment attention to avoid segment fragmentation. These enhancements result in an architecture that outperforms ex-isting SOTA architectures with an average perplexity improvement of 8.5% over similar model sizes.