Cache-to-Cache: Direct Semantic Communication Between Large Language Models

Tianyu Fu, Zihan Min, Hanling Zhang, Jichao Yan, Guohao Dai, Wanli Ouyang, Yu Wang · Oct 3, 2025 · Citations: 0

Abstract

Multi-LLM systems harness the complementary strengths of diverse Large Language Models, achieving performance and efficiency gains that are not attainable by a single model. In existing designs, LLMs communicate through text, forcing internal representations to be transformed into output token sequences. This process both loses rich semantic information and incurs token-by-token generation latency. Motivated by these limitations, we ask: Can LLMs communicate beyond text? Oracle experiments show that enriching the KV-Cache semantics can improve response quality without increasing cache size, supporting KV-Cache as an effective medium for inter-model communication. Thus, we propose Cache-to-Cache (C2C), a new paradigm for direct semantic communication between LLMs. C2C uses a neural network to project and fuse the source model's KV-cache with that of the target model to enable direct semantic transfer. A learnable gating mechanism selects the target layers that benefit from cache communication. Compared with text communication, C2C utilizes the deep, specialized semantics from both models, while avoiding explicit intermediate text generation. Experiments show that C2C achieves 6.4-14.2% higher average accuracy than individual models. It further outperforms the text communication paradigm by approximately 3.1-5.4%, while delivering an average 2.5x speedup in latency. Our code is available at https://github.com/thu-nics/C2C.

HFEPX Relevance Assessment

This paper appears adjacent to HFEPX scope (human-feedback/eval), but does not show strong direct protocol evidence in metadata/abstract.

Eval-Fit Score

0/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Detected

HFEPX Fit

Adjacent candidate

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Human Data Lens

Uses human feedback: No
Feedback types: None
Rater population: Unknown
Unit of annotation: Scalar
Expertise required: Coding
Extraction source: Persisted extraction

Evaluation Lens

Evaluation modes: Automatic Metrics
Agentic eval: None
Quality controls: Not reported
Confidence: 0.35
Flags: low_signal, possible_false_positive

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

accuracylatency

Research Brief

Deterministic synthesis

Thus, we propose Cache-to-Cache (C2C), a new paradigm for direct semantic communication between LLMs. HFEPX signals include Automatic Metrics with confidence 0.35. Updated from current HFEPX corpus.

Generated Mar 5, 2026, 4:11 AM · Grounded in abstract + metadata only

Key Takeaways

Thus, we propose Cache-to-Cache (C2C), a new paradigm for direct semantic communication between LLMs.
Experiments show that C2C achieves 6.4-14.2% higher average accuracy than individual models.
Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Researcher Actions

Treat this as method context, then pivot to protocol-specific HFEPX hubs.
Identify benchmark choices from full text before operationalizing conclusions.
Validate metric comparability (accuracy, latency).

Caveats

Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
Low-signal flag detected: protocol relevance may be indirect.

Recommended Queries

human-eval protocol design pairwise preference data quality inter-rater agreement adjudication

Research Summary

Contribution Summary

Thus, we propose Cache-to-Cache (C2C), a new paradigm for direct semantic communication between LLMs.
Experiments show that C2C achieves 6.4-14.2% higher average accuracy than individual models.
It further outperforms the text communication paradigm by approximately 3.1-5.4%, while delivering an average 2.5x speedup in latency.

Why It Matters For Eval

Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Researcher Checklist

Gap: Human feedback protocol is explicit

No explicit human feedback protocol detected.
Pass: Evaluation mode is explicit

Detected: Automatic Metrics
Gap: Quality control reporting appears

No calibration/adjudication/IAA control explicitly detected.
Gap: Benchmark or dataset anchors are present

No benchmark/dataset anchor extracted from abstract.
Pass: Metric reporting is present

Detected: accuracy, latency

Category-Adjacent Papers (Broader Context)

These papers are nearby in arXiv category and useful for broader context, but not necessarily protocol-matched to this paper.

FlashEvaluator: Expanding Search Space with Parallel Evaluation Category Neighbor

Citations: 0 Relevance: 6.15
- Shared arXiv category (cs.CL, cs.LG)
- Shared metric mentions
- Shared terminology (accuracy, latency, direct)
Cross-Family Speculative Prefill: Training-Free Long-Context Compression with Small Draft Models Category Neighbor

Citations: 0 Relevance: 3.30
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy, semantic)
Efficient Self-Evaluation for Diffusion Language Models via Sequence Regeneration Category Neighbor

Citations: 0 Relevance: 3.30
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy, semantic)
From Variance to Invariance: Qualitative Content Analysis for Narrative Graph Annotation Category Neighbor

Citations: 0 Relevance: 3.20
- Shared arXiv category (cs.CL, cs.LG)
MUSE: A Run-Centric Platform for Multimodal Unified Safety Evaluation of Large Language Models Category Neighbor

Citations: 0 Relevance: 3.20
- Shared arXiv category (cs.CL, cs.LG)
When Numbers Tell Half the Story: Human-Metric Alignment in Topic Model Evaluation Category Neighbor

Citations: 0 Relevance: 3.20
- Shared arXiv category (cs.CL, cs.LG)
According to Me: Long-Term Personalized Referential Memory QA Category Neighbor

Citations: 0 Relevance: 2.85
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy)
AdaPonderLM: Gated Pondering Language Models with Token-Wise Adaptive Depth Category Neighbor

Citations: 0 Relevance: 2.85
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy)

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote