Phi-3 Safety Post-Training: Aligning Language Models with a "Break-Fix" Cycle.
Emman HaiderDaniel Perez-BeckerThomas PortetPiyush MadanAmit GargDavid MajercakWen WenDongwoo KimZiyi YangJianwen ZhangHiteshi SharmaBlake BullwinkelMartin PouliotAmanda MinnichShiven ChawlaSolianna HerreraShahed WarrethMaggie EnglerGary LopezNina ChikanovRaja Sekhar Rao DheekondaBolor-Erdene JagdagdorjRoman LutzRichard LundeenTori WesterhoffPete BryanChristian SeifertRam Shankar Siva KumarAndrew BerkleyAlex KesslerPublished in: CoRR (2024)
Keyphrases
- language model
- language modeling
- n gram
- probabilistic model
- retrieval model
- document retrieval
- information retrieval
- speech recognition
- statistical language models
- test collection
- language modelling
- query expansion
- smoothing methods
- training set
- ad hoc information retrieval
- vector space model
- pseudo relevance feedback
- relevance model
- translation model
- language models for information retrieval
- context sensitive
- language model for information retrieval
- document ranking
- retrieval effectiveness
- query terms
- bayesian networks