Label-flip attacks on log-linear DPO reduce to binary sparse approximation problems that can be solved efficiently by lattice-based and binary matching pursuit methods with recovery guarantees.
arXiv preprint arXiv:2402.01920 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Efficient Preference Poisoning Attack on Offline RLHF
Label-flip attacks on log-linear DPO reduce to binary sparse approximation problems that can be solved efficiently by lattice-based and binary matching pursuit methods with recovery guarantees.