Sign inGet started
Paper · arXiv

EdgeRock: Strong Reasoning at Constant Memory

We present EdgeRock, a family of multimodal models that deliver strong reasoning at constant memory. Most models keep reasoning history in a key-value (KV) cache that grows with every token, consuming accelerator memory, limiting concurrent requests, and competing with perception and control on edge devices. With fixed history, memory is allocated once and never runs out, however long the model reasons. Constant memory alone is not new: bounded-state models (recurrent, state-space, and sliding-window) already achieve it, but often at the cost of reasoning, because they cannot keep the distant details that later reasoning needs. We therefore introduce Rock Memory Anchors (RMA): the model l...

View on arXiv
Report an issue
AuthorsHuishan Ji, Weiming Zhang, Qirui Ye et al.
Primary categorycs.LG
Published7 October 2026
Last updated at source16 October 2026
cs.LGcs.CL
Embed this entry
<iframe src="https://silytics.si/embed/item/658193ad-9a93-4fcb-aa07-17a49b37cab7" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 5 days

0 now. Counts as reported by the source, snapshotted daily by Silytics.

Connected in the knowledge graph

Explore this entry in the graph

Recent changes

  • 20 Oct 2026 Name: YANchor-4B: Effective Long-Horizon Reasoning in O(N) Time with O(1) Memory → EdgeRock: Strong Reasoning at Constant Memory
  • 20 Oct 2026 Description: 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 → We present EdgeRock, a family of multimodal models that deliver strong reasoning at constant memory. Most models keep reasoning history in a key-value (KV) cache that grows with every token, consuming accelerator memory, limiting concurrent

Changes detected by Silytics between refreshes of the source.

Models and datasets referencing this paper

Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 20 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.