Constrained Satisfaction Generation
Constrained satisfaction generation refers to generation tasks in which a language model must satisfy specific hard conditions when producing text, such as forbidden characters, format control, and keyword avoidance. It is used to test a model's ability to comply with rigid rules.
Timeline (last 90 days)
Constrained generation refers to a model having to satisfy specific hard conditions when generating text, such as banning certain words, following a specific format, or controlling sentiment polarity.
When evaluating constrained generation, one usually needs to measure both constraint compliance rate and language quality (such as perplexity and human ratings), and the two often involve a trade-off.
All Facts (2)
Constrained generation refers to a model having to satisfy specific hard conditions when generating text, such as banning certain words, following a specific format, or controlling sentiment polarity.
50%UnverifiedWhen evaluating constrained generation, one usually needs to measure both constraint compliance rate and language quality (such as perplexity and human ratings), and the two often involve a trade-off.
50%