# YANchor-4B: Effective Long-Horizon Reasoning in O(N) Time
Source: https://savrn.com/papers/yanchor-4b-effective-long-horizon-reasoning-in-o-n-time-with-o-1-memory
Markdown alternate of the page above; the site index is https://savrn.com/llms.txt

---

## Abstract

Long-horizon reasoning demands access to earlier information at a manageable generation cost. Full-history attention incurs growing storage and computation, while recurrent compression can lose precise details. Therefore, we present YANchor-4B, a general-purpose recurrent model that preserves crucial memory as ANchors for retrieval during subsequent reasoning. Beyond O(N)-time generation and O(1) memory, YANchor enables effective long-horizon reasoning through its multidimensional memory mechanism. For example, on challenging math problems, it achieves 82.93% mean pass@1 on AIME 2024--2026 and 63.64% on HMMT, substantially outperforming linear-time, constant-state counterparts, including larger models. It also delivers several-fold higher batched long-generation throughput than Transformer and hybrid baselines on H100. Furthermore, evaluations across dozens of benchmarks demonstrate YANchor's superiority in general-purpose capabilities.

[Full paper on arXiv](https://arxiv.org/abs/2610.10118)

## Details

arXiv identifier

2610.10118

Published

2026-10-07

Authors

Huishan Ji, Hua Xu, Weiming Zhang, Qirui Ye

## Open Models Built on This Paper

Every model in the SAVRN Model Hub whose card cites this paper, most downloaded first, with what it takes to run each one.

| Model | Task | Size | License | Monthly downloads | Cheapest setup at 16-bit |
| --- | --- | --- | --- | --- | --- |
| [EdgeRock-4B-GGUF](https://savrn.com/models/edgerock-4b-gguf) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | — | other | 714 | — |
| [EdgeRock-4B-MLX-6bit-g64](https://savrn.com/models/edgerock-4b-mlx-6bit-g64) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 412 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [EdgeRock-4B-ONNX-Q4-CPU-g32](https://savrn.com/models/edgerock-4b-onnx-q4-cpu-g32) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | — | other | 381 | — |
| [EdgeRock-4B-ONNX-Q4-g32](https://savrn.com/models/edgerock-4b-onnx-q4-g32) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | — | other | 372 | — |
| [EdgeRock-4B-ONNX-Q8](https://savrn.com/models/edgerock-4b-onnx-q8) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | — | other | 371 | — |
| [EdgeRock-4B-MLX-4bit-g32](https://savrn.com/models/edgerock-4b-mlx-4bit-g32) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 62 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [EdgeRock-4B-MLX-8bit](https://savrn.com/models/edgerock-4b-mlx-8bit) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 60 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [EdgeRock-4B-MLX-4bit-mixed](https://savrn.com/models/edgerock-4b-mlx-4bit-mixed) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 55 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [EdgeRock-4B-AWQ](https://savrn.com/models/edgerock-4b-awq) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 32 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [EdgeRock-4B-GPTQ](https://savrn.com/models/edgerock-4b-gptq) Aaron Xu | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | 28 | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |
| [YANchor-4B](https://savrn.com/models/yanchor-4b) RocoreMatrix | [Text generation](https://savrn.com/models/tasks/text-generation) | 4.6B | other | — | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) $1.85/hr |

## Explore More

- [All research papers](https://savrn.com/papers)
- [The model directory](https://savrn.com/models)
- [Datasets](https://savrn.com/datasets)

## Source

- arXiv metadata, read 2026-10-23.
- [How the hub is built](https://savrn.com/model-hub/methodology)