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

Simon P. Villani

Sep 2026 · 0 citations · 24 references

Computer Science

TL;DR

This work demonstrates useful persistent hybrid-state transfer across one architecture-matched Qwen3.5 4B-to-9B sibling pair, the first demonstrated cross-model handoff of persistent recurrent inference state between differently sized hybrid language models without target prefix replay.

Abstract

Can one language model hand its live memory to another without the receiver rereading the context? We demonstrate useful persistent hybrid-state transfer across one architecture-matched Qwen3.5 4B-to-9B sibling pair. To our knowledge, this is the first demonstrated cross-model handoff of persistent recurrent inference state between differently sized hybrid language models without target prefix replay. Translated attention KV alone leaves a large gap; adding the Gated DeltaNet (GDN) persistent-state package lowers teacher-forced negative log-likelihood (NLL), the average next-token log-loss, by 0.747 nats/token (95% paired document bootstrap CI [0.6921, 0.8047]), improving all 64 PG19 documents. Direct recurrent and convolution reuse outperforms the tested learned GDN maps, consistent with partial functional compatibility of persistent-state coordinates. A fresh component factorial selects translated KV with direct recurrent and convolution state. An additional 434,176-parameter correction improves that base on 64 fresh web documents: continuation loss is 0.076 nats/token above native 9B (excess NLL), Jensen-Shannon (JS) divergence is 0.022, and native context recovery (NCR) is 0.918. Corrected 9B significantly beats continued 4B inference while processing zero historical prefix tokens. Evidence covers one direction, one geometry-matched Base-model pair, and 4K teacher-forced continuation; the near-native gate failed, the 16K branch was not run, and free-generation equivalence and a general state interface remain unproven.

DOI arXiv Semantic Scholar

View source

Similar papers

#natural language process... Preprint Sep 2026

What Attention Recalls and Recurrence Controls in Hybrid Language Models

Two cache-level interventions are introduced: Split-prefill keeps only the KV cache or only the recurrent state from a prefilled context, then generates an answer, and state-swap pairs the KV cache from one context with the recurrent state from another in a single forward pass.

Kirill Afendulev, Alexey Dontsov, Elena Tutubalina et al. · 2 citations · ⚡1

#natural language process... Review Sep 2026

The Evolution of Attention in Large Language Models: Mechanisms, Trade-offs, and Emerging Trends

Self-attention gives LLMs fine-grained, query-dependent access to context, but dense token interactions incur quadratic prefill cost and a key--value cache growing with context length. Research thus spans explicit-memory compression, sparse access, recurrent state construction, structured state dynamics, and heterogene...

Zhen-Tao Tan, Jing-Yi Shen, Yan-Bo Li et al. · 0 citations

#natural language process... Preprint Sep 2026

Lngram v2: Latent N-Gram Memory with Interpretable Discrete Representations

Analysis shows that Lngram v2 substantially reduces both total and activated memory parameters while maintaining or improving language modeling performance, and shows that its discrete IDs preserve substantial semantic structure of continuous hidden states, enabling semantic recovery from IDs alone and stable ID--seman...

Yun Zheng, Bin Wen, Xiao-Jie Wang et al. · 0 citations

Preprint Aug 2026

Dual-Cache Latent Space Communication between Heterogeneous Language Models

XKV is introduced, which lifts all three restrictions on multi-agent LLM answering: learned-query attention pools both caches; self-attention over receiver-aligned layer tokens, with a learned layer map reconciling different depths, mixes the pooled summaries into a compact joint memory.

Ji-Yao Liu, Qi Zhang, Yaoyi Jia et al. · 3 citations

#artificial intelligence Preprint Oct 2026

Hybrid Latent Attention for Looped Language Models

Looped language models apply the same stack of layers T times to each token, which deepens the model without adding parameters but multiplies its key-value (KV) cache by T. The larger cache limits how many sequences a GPU can decode at once and slows each decoding step, which reads the whole cache. We propose Hybrid La...

Yu-Han Chen, Si-Yuan Zhang, Nan Wang et al. · 0 citations

Preprint Aug 2026

Architecture-Dependent Causal Transfer of Activation States Across Large Language Models

End-to-end activation-state transfer between LLMs, as currently implemented, is architecture-dependent rather than universal, and it is concluded that end-to-end activation-state transfer between LLMs is architecture-dependent rather than universal.

Fernando Cardenas Piepereit · 3 citations