Language models (LMs) often generate incoherent outputs: they refer to events and entity states that are incompatible with the state of the world described in their inputs. We introduce SituationSupervision, a family of approaches for improving coherence in LMs by training them to construct and condition on explicit representations of entities and their states. SituationSupervision has two components: an auxiliary situation modeling task that trains models to predict state representations in context, and a latent state inference procedure that imputes these states from partially annotated training data. SituationSupervision can be applied to both fine-tuning (by supervising LMs to encode state variables in their hidden representations) and prompting (by inducing LMs to interleave textual descriptions of entity states with output text). In both cases, SituationSupervision requires only a small number of state annotations to produce major coherence improvements (between 4-11%), showing that standard LMs can be sample-efficiently trained to model not just language but the situations it describes.
翻译:语言模型(LMS)往往产生不一致的产出:它们指的是与投入中描述的世界状况不相符的事件和实体国家。我们引入了“情况监视”这一提高LMS一致性的一套方法,通过培训它们建立实体及其州的明确代表和条件来提高LMS的一致性。情况监视有两个组成部分:一个辅助情况模拟任务,培训模型以预测在背景中的国家代表情况,一个潜在状态推断程序,将这些国家从部分附加说明的培训数据中估算出来。情况监视可适用于微调(监督LMS在其隐藏的表述中对变量进行编码)和快速调整(引导LMS用输出文本对实体国家进行截断文字描述 ) 。 在这两种情况下,情况监视只需要少量的状态说明就可以产生重大的一致性改进(4-11%之间),表明标准LMS程序可以经过抽样有效的培训,不仅模拟语言,而且模拟其描述的情况。