When can transformers compositionally generalize in-context?
Many tasks can be composed from a few independent components. This gives rise to a combinatorial explosion of possible tasks, only some of which might be encountered during training. Under what circumstances can transformers compositionally generalize from a subset of tasks to all possible combinations of tasks that share similar components? Here we study a modular multitask setting that allows us to precisely control compositional structure in the data generation process. We present evidence that transformers learning in-context struggle to generalize compositionally on this task despite being in principle expressive enough to do so. Compositional generalization becomes possible only when introducing a bottleneck that enforces an explicit separation between task inference and task execution.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
When Can Transformers Ground and Compose: Insights from Compositional Generalization Benchmarks
Humans can reason compositionally whilst grounding language utterances to the real world. Recent benchmarks like ReaSCAN use navigation tasks grounded in a grid world to assess whether neural models exhibit similar capab…
Compositional Generalization (AVG)The Impact of Depth on Compositional Generalization in Transformer Language Models
To process novel sentences, language models (LMs) must generalize compositionally -- combine familiar elements in new ways. What aspects of a model's structure promote compositional generalization? Focusing on transforme…
Language ModelingLanguage ModellingMaking Transformers Solve Compositional Tasks
Several studies have reported the inability of Transformer models to generalize compositionally, a key type of generalization in many NLP tasks such as semantic parsing. In this paper we explore the design space of Trans…
Semantic ParsingMaking Transformers Solve Compositional Tasks
Several studies have reported the inability of Transformer models to generalize compositionally, a key type of generalization in many NLP tasks such as semantic parsing. In this paper we explore the design space of Trans…
Semantic ParsingThe Coverage Principle: A Framework for Understanding Compositional Generalization
Large language models excel at pattern matching, yet often fall short in systematic compositional generalization. We propose the coverage principle: a data-centric framework showing that models relying primarily on patte…