arXiv Machine Learning July 23, 2026 By Keston Aquino-Michaels

The Orthogonalized Read Is a Removable Training Scaffold for Recurrent Memory

Read the original on arXiv Machine Learning →

arXiv:2607. 19390v1 Announce Type: new Abstract: A recent report finds that orthogonalizing the mLSTM memory matrix at read time (five Newton-Schulz iterations, trained through) substantially improves noisy associative recall.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

benchmarks

arXiv Machine Learning

Jul 24

Memoir: Should a Model Write to Its Memory While It Thinks?

arXiv:2607. 20792v1 Announce Type: new Abstract: Memoir combines per-sample fast memory, shared slow parameters, variable-depth latent recurrence, and a future-latent energy objective.

By Jaber Jaber, Osama Jaber

arXiv AI

Jul 24

Naju: A Native Discrete State-Space Model with Independent Retention and Writing for Long-Sequence Memory

arXiv:2607. 21000v1 Announce Type: new Abstract: Long-sequence memory tracking places two opposing demands on a recurrent state: near-lossless retention of stored bindings over long horizons, and active overwriting of stale ones.

By Hyuk Lim, Seunghyun Yoon

llms

Hugging Face Trending Papers

Jul 23

Naju: A Native Discrete State-Space Model with Independent Retention and Writing for Long-Sequence Memory

Long-sequence memory tracking places two opposing demands on a recurrent state: near-lossless retention of stored bindings over long horizons, and active overwriting of stale ones. In our diagnostic suite, the strongest efficient baselines tend to solve only one side well.

llms

arXiv Machine Learning

2d ago

Finite-Horizon Fisher Memory in Two-Sided Power-Bounded Recurrent Systems

arXiv:2609.39800v1 Announce Type: new Abstract: We analyse allocation, admission and post-write retention in finite-horizon linear-Gaussian noisy recurrent memories. At every horizon, the directional...

By Jeonghoon Lee (Attractor Dynamics Inc.)

arXiv Machine Learning

Aug 14

Gradient-Free Warm-Start Library Recovery: an Amortized-Regret Separation

arXiv:2606. 21253v2 Announce Type: replace Abstract: Continual learning that is gradient-free, local, online, and append-only is attractive for edge and streaming deployment, but its value is usually argued informally.

By Jianwei Lou (RailMind Systems, Neuss, Germany)

agents computer-vision

arXiv AI

Aug 25

Crossing the Margin Cliff: Toward Relearn-Robust LLM Unlearning via Margin Calibration

arXiv:2607.27836v2 Announce Type: replace Abstract: Large language model unlearning is consistently fragile under relearn attacks. On TOFU, fine-tuning on twenty forget examples substantially recover...

By Xiangyu Yin, Jiaxu Liu, Zhen Chen, Chih-Hong Cheng

llms fine-tuning efficiency

One email a morning, machine-written

Related stories

Memoir: Should a Model Write to Its Memory While It Thinks?

Naju: A Native Discrete State-Space Model with Independent Retention and Writing for Long-Sequence Memory

Naju: A Native Discrete State-Space Model with Independent Retention and Writing for Long-Sequence Memory

Finite-Horizon Fisher Memory in Two-Sided Power-Bounded Recurrent Systems

Gradient-Free Warm-Start Library Recovery: an Amortized-Regret Separation

Crossing the Margin Cliff: Toward Relearn-Robust LLM Unlearning via Margin Calibration