Search papers, labs, and topics across Lattice.
This study investigates the privacy implications of sparse collaborative inference (CI), where model computations are split between edge devices and servers. By analyzing the transmission of sparse activations and their positional information, the authors reveal that while sparsification reduces communication costs, it does not significantly mitigate privacy risks, as the positions of retained values can lead to high-fidelity input reconstructions. The findings highlight that the positional data of sparse activations poses a substantial privacy threat, necessitating careful auditing in CI systems.
Sparsifying activations in collaborative inference may cut costs, but it exposes a hidden privacy risk from the positions of those activations that could enable re-identification.
Collaborative inference (CI) splits a model between an edge device and a server, whereby the client computes an intermediate activation, transmits it, and the server completes the computation. This raises two concerns, the communication cost of the transmission and the risk that it reveals private information about the input. Recent work reduces this cost by sparsifying activations and entropy-coding the result. Sparsity has also been argued to improve privacy, on the intuition that transmitting fewer values reveals less about the input. We test this claim by decomposing the sparse activation into the retained values and the set of positions they occupy, and by reconstructing inputs from each component in isolation. We find that sparsification reduces the leakage far less than it reduces the transmission cost, and that the remaining risk shifts to the positions, which prior analyses treat as side information for decoding. Across natural-image and face datasets, the positions alone constitute a serious privacy risk, enabling high-fidelity reconstructions and re-identification of individuals. The leakage from the positions persists even when both the transmission cost and the task utility are low. We conclude that the positions of sparse activations should be treated as sensitive transmitted data and audited carefully in the context of collaborative inference. Code is available at https://github.com/an7123/Privacy-Study-Sparse-CI.
