LLM Digest
Subscribe

Story

arxiv_cs_lg · Nov 18, 2026 · paper

Source brief

Stealing Exact Training Data from Other Clients during Federated Learning

arxiv.orgNov 18, 2026
original source linked

In brief

Malicious adversaries can employ gradient inversion attacks to reverse-engineer private data using the model updates of a round in Federated Learning. While there have been developments in gradient inversion for text, exact text recovery by a malicious client has so far remained

Continue reading

Read the original at arxiv.org →Open in live feedRead that day’s brief

Earlier in this thread 4 items