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

arXiv:2609.25053 · cs.CL, cs.AI · Submitted 2026-09-07 · Read on arXiv

cs.CL, cs.AI

Submitted: 2026-09-07

Updated: 2026-09-07

Comments: 14 pages, 6 figures, 12 tables

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• KV cache — 6× in this paper · explained in 3 episodes
• Accuracy — 5× in this paper · explained in 1 episode
• Adaptation — 5× in this paper · explained in 1 episode
• Calibration — 5× in this paper · explained in 7 episodes
• Fidelity — 5× in this paper · explained in 2 episodes
• Reconstruction Error — 5× in this paper · explained in 2 episodes
• Ablation — 4× in this paper · explained in 1 episode
• Alignment — 4× in this paper · explained in 1 episode
• Cross-Model Transfer — 3× in this paper · explained in 2 episodes
• Verification — 3× in this paper · explained in 2 episodes
• Distillation — 2× in this paper · explained in 2 episodes
• Frozen model — 2× in this paper · explained in 3 episodes
• Regularization — 2× in this paper · explained in 3 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Aggregation — 1× in this paper · explained in 1 episode
• Alignment Scores — 1× in this paper · explained in 1 episode
• Benchmark — 1× in this paper · explained in 2 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Latent Space — 1× in this paper · explained in 7 episodes
• Quantization — 1× in this paper · explained in 12 episodes
• Trace — 1× in this paper · explained in 2 episodes

Sources

• Cross-Model KV Cache Transfer in LLM Families: A Closed-Form Linear Mapping for Prefill Reuse
• KV Cache Translation across Heterogeneous Large Language Models
• Frozen Memory Is Not Enough: Rethinking External Memory as Extraction
• A Universal Context-Reuse Layer for Cross-Model KV Sharing
• HYPIC: Accelerating Hybrid-Attention LLM Serving with Position-Independent Caching
• Tail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLMs
• Semantic Cache Distillation: Efficient State Transfer via Reuse and Selective Patching
• Marconi: Prefix Caching for the Era of Hybrid LLMs
• The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale
• StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems
• CacheBridge: Efficient Cross-Model KV Cache Transfer
• Compressive Transformers for Long-Range Sequence Modelling
• Latent Cache Flow: Model-to-Model Communication Without Text
• Gated Delta Networks: Improving Mamba2 with Delta Rule
• S0 Tuning: Zero-Overhead Adaptation of Hybrid Recurrent-Attention Models
• WriteSAE: Sparse Autoencoders for Recurrent State
• DASC: Decay-Aware State Compression for Hybrid Linear-Attention Serving
• Metis: Memory Foundation Model

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering