Guardado en:
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2509.24257 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909997341868032 |
|---|---|
| author | Wang, Ke Zhao, Zishuo Song, Xinyuan Li, Zelin Xia, Libin Tong, Chris Shi, Bill Qu, Wenjie Yang, Eric Ai, Lynn |
| author_facet | Wang, Ke Zhao, Zishuo Song, Xinyuan Li, Zelin Xia, Libin Tong, Chris Shi, Bill Qu, Wenjie Yang, Eric Ai, Lynn |
| contents | Decentralized inference provides a scalable and resilient paradigm for serving large language models (LLMs), enabling fragmented global resource utilization and reducing reliance on centralized providers. However, in a permissionless environment without trusted nodes, ensuring the correctness of model outputs remains a core challenge. We introduce VeriLLM, a publicly verifiable protocol for decentralized LLM inference that achieves security with incentive guarantees while maintaining practical efficiency. VeriLLM combines lightweight empirical rerunning with minimal on-chain checks to preclude free-riding, allowing verifiers to validate results at approximately 1% of the underlying inference cost by exploiting the structural separation between prefill and autoregressive decoding. To prevent verification bottlenecks, we design an isomorphic inference--verification architecture that multiplexes both inference and verification roles across the same GPU workers. This design (i) improves GPU utilization and overall throughput, (ii) enlarges the effective validator set, enhancing robustness and liveness, and (iii) enforces task indistinguishability to prevent node-specific optimizations or selective behavior. Through theoretical analysis and system-level evaluation, we show that VeriLLM achieves reliable public verifiability with minimal overhead, offering a practical foundation for trustworthy and scalable decentralized LLM inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_24257 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VeriLLM: A Lightweight Framework for Publicly Verifiable Decentralized Inference Wang, Ke Zhao, Zishuo Song, Xinyuan Li, Zelin Xia, Libin Tong, Chris Shi, Bill Qu, Wenjie Yang, Eric Ai, Lynn Cryptography and Security Machine Learning C.2.1 Decentralized inference provides a scalable and resilient paradigm for serving large language models (LLMs), enabling fragmented global resource utilization and reducing reliance on centralized providers. However, in a permissionless environment without trusted nodes, ensuring the correctness of model outputs remains a core challenge. We introduce VeriLLM, a publicly verifiable protocol for decentralized LLM inference that achieves security with incentive guarantees while maintaining practical efficiency. VeriLLM combines lightweight empirical rerunning with minimal on-chain checks to preclude free-riding, allowing verifiers to validate results at approximately 1% of the underlying inference cost by exploiting the structural separation between prefill and autoregressive decoding. To prevent verification bottlenecks, we design an isomorphic inference--verification architecture that multiplexes both inference and verification roles across the same GPU workers. This design (i) improves GPU utilization and overall throughput, (ii) enlarges the effective validator set, enhancing robustness and liveness, and (iii) enforces task indistinguishability to prevent node-specific optimizations or selective behavior. Through theoretical analysis and system-level evaluation, we show that VeriLLM achieves reliable public verifiability with minimal overhead, offering a practical foundation for trustworthy and scalable decentralized LLM inference. |
| title | VeriLLM: A Lightweight Framework for Publicly Verifiable Decentralized Inference |
| topic | Cryptography and Security Machine Learning C.2.1 |
| url | https://arxiv.org/abs/2509.24257 |