arXiv Computation and Language September 1, 2026 By Jiayu Hou, Lei Wang

When to Adapt: Conditional Memory Adapters for Retention-Preserving Domain Specialization

Read the original on arXiv Computation and Language →

The paper introduces Engram Adapter, a conditional memory adapter that selectively activates on in-domain inputs while suppressing out-of-domain (OOD) retrievals. It leverages multi‑channel matching over local n‑gram patterns with occupancy tracking and a learned scalar gate to inject residuals only when appropriate. Experiments on Qwen3-4B/8B with AG‑News and MedMCQA show improved in‑domain accuracy while preserving nearly all OOD performance, outperforming always‑on baselines on LegalBench.

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 Computation and Language.

arXiv Computation and Language

Sep 23

LatentPort: Beyond KV Cache - Cross-Model Transfer of Recurrent Memory in Hybrid Language Models: A 4B-to-9B Hybrid-State Handoff Without Target Prefix Replay

The paper introduces LatentPort, a method that allows a language model to transfer its live memory to another model without requiring the receiver to reread the context. Experiments on a Qwen3.5 4B-to-9B sibling pair show that adding a Gated DeltaNet (GDN) persistent-state package reduces negative log‑likelihood by 0.747 nats/token and improves performance across 64 PG19 documents. The study also demonstrates that direct recurrent and convolution reuse outperforms learned GDN maps, and a 434,176‑parameter correction further narrows the performance gap to the native 9B model.

By Simon P. Villani

llms efficiency

arXiv Computation and Language

4d ago

ATTUNER: Recomputation-Free KV Cache Reuse via Query-Side Adaptation

arXiv:2609.36722v1 Announce Type: new Abstract: Large language model (LLM) agents repeatedly load reusable content, such as skills, documents, and memory entries, into the current context. Re-encodin...

By Xinghao Chen, Junnan Dong, Cai Ke, Chak Tou Leong, Haocheng Sun, Keyu Chen, Siyu An, Ruizhi Qiao, Xing Sun, Wenjie Li, Xiaoyu Shen

llms agents efficiency benchmarks

arXiv Machine Learning

Jul 22

A Better Start for Language Models: Domain-Conditional Position Offsets

arXiv:2607. 18302v1 Announce Type: new Abstract: Autoregressive language models are least accurate at the beginning of a sequence, where little context forces reliance on a generic pretraining prior.

By Ye Qiao

llms rag fine-tuning safety

arXiv AI

Jun 10

Null-Space Constrained Low-Rank Adaptation for Response-Specified Large Language Model Unlearning

arXiv:2606. 10989v1 Announce Type: new Abstract: Large language model unlearning aims to suppress designated undesirable knowledge while preserving benign capabilities.

By Bocheng Ju, Jianhua Wang, Chengliang Liu, Xiaolin Chang

llms fine-tuning benchmarks safety

arXiv Machine Learning

Aug 24

Nothing Changed but the Model: CellFill -- Bounded In-Cell Learning for Bit-Identical, Revocable Updates to Quantized LLMs

arXiv:2608.20873v1 Announce Type: new Abstract: Every way of teaching a deployed language model something new -- full fine-tuning, adapter merging, model editing -- replaces the released checkpoint,...

By Zifeng Liu, Zhiyong Du, Yaxin Lu, Yiming Mao, Zhenhe Wang, Wenqi Shi, Zhengkun Jing

llms fine-tuning efficiency

arXiv Machine Learning

Aug 7

QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding

arXiv:2608. 05326v1 Announce Type: new Abstract: Autoregressive large language model inference is increasingly constrained by the memory footprint of the Key-Value (KV) cache.

By Ayushman Garg, Akshita Gupta, Shaswata Bhattacharya, Abhishek Gupta, Sandeep Kumar, Manoj Kumar

llms efficiency benchmarks

One email a morning, machine-written

Related stories

LatentPort: Beyond KV Cache - Cross-Model Transfer of Recurrent Memory in Hybrid Language Models: A 4B-to-9B Hybrid-State Handoff Without Target Prefix Replay

ATTUNER: Recomputation-Free KV Cache Reuse via Query-Side Adaptation

A Better Start for Language Models: Domain-Conditional Position Offsets

Null-Space Constrained Low-Rank Adaptation for Response-Specified Large Language Model Unlearning

Nothing Changed but the Model: CellFill -- Bounded In-Cell Learning for Bit-Identical, Revocable Updates to Quantized LLMs

QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding