TL++: Accuracy and Privacy Preserving Traversal Learning for Distributed Intelligent Systems
Not on the map. Why?
Abstract (arXiv)
Distributed intelligent systems increasingly need to train across data silos without centralizing raw data. Federated learning keeps data local but can suffer under heterogeneous partitions and requires repeated full-model exchange. Split learning reduces communication through cut-layer activations, but standard protocols generally do not recover centralized mini-batch gradient behavior and may expose activations and gradients in plaintext. We present TL++, a two-mode traversal-learning framework that constructs virtual batches across nodes to recover centralized mini-batch gradient behavior under explicit synchronization assumptions. Base mode exchanges cut-layer activations and gradients rather than full models. Secure mode secret-shares each cut-layer activation and gradient between an orchestrator and a non-colluding helper, preventing either server from observing plaintext cut-layer tensors. This protection is limited to a semi-honest two-server setting; labels and loss-related outputs remain visible to the orchestrator. In the lightweight secure path evaluated here, exactness requires a linear or affine server path, while nonlinear operations require nonlinear MPC or approximation. We formalize TL++, analyze communication and computation costs, and evaluate it against federated and split-learning baselines on CIFAR-10 and BioGPT/PubMedQA using full fine-tuning and LoRA. On CIFAR-10, TL++ base cut 1 and exact secure cut 3 achieve accuracies of 91.41% (SD 0.19) and 90.93% (SD 0.17), respectively, exceeding the strongest measured non-TL++ baseline by more than 12 percentage points. TL++ base cut 1 also reduces per-step communication by 13.1-fold relative to full-model synchronization. PubMedQA results similarly favor TL++. Overall, TL++ approaches centralized-training performance while reducing communication and providing activation-level secret sharing.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 40 of its 66 references that are in the map. First posted on arXiv 24 Jun 2026.
Builds on
40 of 66 references.
References · 40
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
- 2019Parameter-Efficient Transfer Learning for NLPcited by 4,236
- 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,418
- 2019Advances and Open Problems in Federated Learningcited by 3,046
- 2018Federated Optimization in Heterogeneous Networkscited by 2,987
- 2016Deep Learning with Differential Privacycited by 2,728
- 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
- 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,492
- 2016Federated Learning: Strategies for Improving Communication Efficiencycited by 1,475
- 2019PubMedQA: A Dataset for Biomedical Research Question Answeringcited by 1,288
- 2018Federated Learning with Non-IID Datacited by 1,108
28 more
- 2020Adaptive Federated Optimizationcited by 1,046
- 2019Deep Leakage from Gradientscited by 1,021
- 2020Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimizationcited by 906
- 2021Model-Contrastive Federated Learningcited by 766
- 2020Inverting Gradients -- How easy is it to break privacy in federated learning?cited by 712
- 2020Ditto: Fair and Robust Federated Learning Through Personalizationcited by 672
- 2017Renyi Differential Privacycited by 650
- 2018Local SGD Converges Fast and Communicates Littlecited by 605
- 2018signSGD: Compressed Optimisation for Non-Convex Problemscited by 571
- 2021FedBN: Federated Learning on Non-IID Features via Local Batch Normalizationcited by 571
- 2021Federated Learning Based on Dynamic Regularizationcited by 562
- 2022BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Miningcited by 538
- 2019Clustered Federated Learning: Model-Agnostic Distributed Multi-Task Optimization under Privacy Constraintscited by 503
- 2017Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Trainingcited by 478
- 2018Split learning for health: Distributed deep learning without sharing raw patient datacited by 395
- 2020iDLG: Improved Deep Leakage from Gradientscited by 382
- 2019FedPAQ: A Communication-Efficient Federated Learning Method with Periodic Averaging and Quantizationcited by 356
- 2020SplitFed: When Federated Learning Meets Split Learningcited by 310
- 2018Distributed learning of deep neural network over multiple agentscited by 279
- 2019Asynchronous Federated Optimizationcited by 267
- 2022The Forward-Forward Algorithm: Some Preliminary Investigationscited by 266
- 2016Direct Feedback Alignment Provides Learning in Deep Neural Networkscited by 240
- 2017Gradient Sparsification for Communication-Efficient Distributed Optimizationcited by 201
- 2018Gazelle: A Low Latency Framework for Secure Neural Network Inferencecited by 193
- 2020Unleashing the Tiger: Inference Attacks on Split Learningcited by 105
- 2020NoPeek: Information leakage reduction to share activations in distributed deep learningcited by 66
- 2019CrypTFlow: Secure TensorFlow Inferencecited by 61
- 2024CURE: Privacy-Preserving Split Learning Done Rightcited by 6
Built on by
Citing papers per year: 2026: 1.
1 paper.