Pith. sign in

REVIEW 7 cited by

Split-and-Denoise: Protect large language model inference with local differential privacy

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2310.09130 v4 pith:ML6U44AB submitted 2023-10-13 cs.AI cs.CR

classification cs.AIcs.CR
keywords modelprivacydownstreamembeddingsinferencelanguagetasksclient
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Large Language Models (LLMs) excel in natural language understanding by capturing hidden semantics in vector space. This process enriches the value of text embeddings for various downstream tasks, thereby fostering the Embedding-as-a-Service (EaaS) business model. However, the risk of privacy leakage due to direct text transmission to servers remains a critical concern. To address this, we introduce Split-N-Denoise (SnD), an private inference framework that splits the model to execute the token embedding layer on the client side at minimal computational cost. This allows the client to introduce noise prior to transmitting the embeddings to the server, and subsequently receive and denoise the perturbed output embeddings for downstream tasks. Our approach is designed for the inference stage of LLMs and requires no modifications to the model parameters. Extensive experiments demonstrate SnD's effectiveness in optimizing the privacy-utility tradeoff across various LLM architectures and diverse downstream tasks. The results reveal an improvement in performance under the same privacy budget compared to the baselines by over 10\% on average, offering clients a privacy-preserving solution for local privacy protection.

Discussion (0). Sign in to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Learning Obfuscations Of LLM Embedding Sequences: Stained Glass Transform

    cs.LG 2025-06 conditional novelty 6.0 of 10

    A learned stochastic embedding obfuscator, the Stained Glass Transform, is claimed to reduce mutual information between prompts and their server-side representations while preserving LLM utility.

  2. Efficient and Privacy Aware Edge Cloud Collaborative Inference for Large Language Models

    cs.CR 2026-07 conditional novelty 5.0 of 10

    An edge-cloud LLM serving scheme splits the final projection along the hidden dimension, keeps cache control and LoRA modules on-device, and reports 29-46% lower per-token latency and 56-67% lower downlink payloads th...

  3. Enhancing Model Privacy in Federated Learning with Random Masking and Quantization

    cs.LG 2025-08 reject novelty 5.0 of 10

    FedQSN hides part of the server model with random masks and quantizes the remainder to give clients a degraded proxy, reporting a large global-vs-proxy performance gap with modest loss in the final global model.

  4. ISACL: Internal State Analyzer for Copyrighted Training Data Leakage

    cs.CL 2025-08 conditional novelty 5.0 of 10

    An MLP trained on LLM internal states predicts Rouge-L-defined literal copying leakage with high accuracy, but not paraphrase-level leakage.

  5. LLM Access Shield: Domain-Specific LLM Framework for Privacy Policy Compliance

    cs.CR 2025-05 conditional novelty 5.0 of 10

    An enterprise proxy that detects sensitive data in LLM prompts with a fine-tuned small model and replaces it with format-preserving encryption.

  6. Toward Edge General Intelligence with Multiple-Large Language Model (Multi-LLM): Architecture, Trust, and Orchestration

    cs.NI 2025-07 conditional novelty 4.0 of 10

    A survey of multi-LLM systems in edge computing, covering architectures, enabling technologies, trust mechanisms, applications, and open datasets for edge general intelligence.

  7. SoK: Semantic Privacy in Large Language Models

    cs.CR 2025-06 conditional novelty 4.0 of 10

    A systematization of knowledge arguing that LLM privacy threats extend beyond data leakage to semantically inferred attributes, and that current defenses only partially address them.

Pith tools