Rarely a problem? Language models exhibit inverse scaling in their predictions following few-type quantifiers
How well do language models deal with quantification? In this study, we focus on 'few'-type quantifiers, as in 'few children like toys', which might pose a particular challenge for language models because the sentence components with out the quantifier are likely to co-occur, and 'few'-type quantifiers are rare. We present 960 English sentence stimuli from two human neurolinguistic experiments to 22 autoregressive transformer models of differing sizes. Not only do all the models perform poorly on 'few'-type quantifiers, but overall the larger the model, the worse its performance. This inverse scaling is consistent with previous work suggesting that larger models increasingly reflect online rather than offline human processing, and we argue that the decreasing performance of larger models may challenge uses of language models as the basis for natural language systems.
Code (0)
등록된 구현이 없습니다.
Tasks
SentenceVocal Bursts Type PredictionSimilar Papers 제목 키워드 기반
Beyond Positive Scaling: How Negation Impacts Scaling Trends of Language Models
Language models have been shown to exhibit positive scaling, where performance improves as models are scaled up in terms of size, compute, or data. In this work, we introduce NeQA, a dataset consisting of questions with …
NegationQuestion AnsweringTask 2Inverse scaling can become U-shaped
Scaling up language models has been empirically shown to improve performance on a wide range of downstream tasks. However, if we were to observe worse performance as a function of scale ("inverse scaling") on certain tas…
AttributeUncovering Scaling Laws for Large Language Models via Inverse Problems
Large Language Models (LLMs) are large-scale pretrained models that have achieved remarkable success across diverse domains. These successes have been driven by unprecedented complexity and scale in both data and computa…
U-shaped and Inverted-U Scaling behind Emergent Abilities of Large Language Models
Large language models (LLMs) have been shown to exhibit emergent abilities in some downstream tasks, where performance seems to stagnate at first and then improve sharply and unpredictably with scale beyond a threshold. …
Recent scaling properties of Bitcoin price returns
While relevant stylized facts are observed for Bitcoin markets, we find a distinct property for the scaling behavior of the cumulative return distribution. For various assets, the tail index $\mu$ of the cumulative retur…
Time SeriesTime Series Analysis