# Cartesia — Based: Simple linear attention language models balance the recall‑throughput tradeoff

- Company: Cartesia (cartesia.ai)
- Announced: 2024-03-04
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.cartesia.ai/blog/based
- Record: https://forck.live/items/8032-based-simple-linear-attention-language-models-balance-the-recall-throughput
- Subject: Sonic / Ink
- Models affected: Based

Cartesia presents Based, a recurrent architecture that uses sliding window attention and linear attention to outperform prior sub-quadratic models on recall-intensive tasks while achieving faster generation speeds than FlashAttention-2 and Mamba.

## Evidence

Verbatim from https://www.cartesia.ai/blog/based:

> We’re excited to share the latest progress in this line of work. Our recent work digs deeper into the recall challenge. We begin by illustrating a fundamental tradeoff between a model’s recall abilities and its memory consumption during generation. This analysis informs the design of Based, a simple recurrent architecture that outperforms prior sub-quadratic models on real-world recall-intensive tasks (information extraction, reading comprehension) and in-context learning. At the same time, Based offers fast generation speeds: Based is 56% and 44% faster at processing prompts than FlashAttention-2 and Mamba respectively. Based achieves 24x higher text generation throughput than FlashAttention-2.

---

Record: https://forck.live/items/8032-based-simple-linear-attention-language-models-balance-the-recall-throughput
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md