Baseline Defenses for Adversarial Attacks Against Aligned Language Models.
Neel JainAvi SchwarzschildYuxin WenGowthami SomepalliJohn KirchenbauerPing-yeh ChiangMicah GoldblumAniruddha SahaJonas GeipingTom GoldsteinPublished in: CoRR (2023)
Keyphrases
- context sensitive
- language model
- denial of service attacks
- trec test collections
- language modeling
- word error rate
- denial of service
- dos attacks
- n gram
- okapi bm
- information retrieval
- speech recognition
- language modelling
- probabilistic model
- document retrieval
- query expansion
- retrieval model
- spam filters
- smoothing methods
- network security
- test collection
- document length
- vector space model
- query terms
- ad hoc information retrieval
- statistical language models
- ddos attacks
- trec collections
- pseudo relevance feedback
- query specific
- document ranking
- relevance model
- language models for information retrieval