paper-with-me

Papers

When can transformers compositionally generalize in-context?

2024-07-17 · Seijin Kobayashi, Simon Schug, Yassir Akram, Florian Redhardt, Johannes von Oswald, Razvan Pascanu, Guillaume Lajoie, João Sacramento

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.

📄 PDF Abstract BibTeX arXiv:2407.12275

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

When Can Transformers Ground and Compose: Insights from Compositional Generalization Benchmarks

2022-10-23 · Ankur Sikarwar, Arkil Patel, Navin Goyal

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

2023-10-30 · Jackson Petty, Sjoerd van Steenkiste, Ishita Dasgupta, Fei Sha 외

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 Modelling

Making Transformers Solve Compositional Tasks

2021-11-16 · ACL ARR September 2021 9 · Anonymous

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 Parsing

Making Transformers Solve Compositional Tasks

2021-08-09 · ACL 2022 5 · Santiago Ontañón, Joshua Ainslie, Vaclav Cvicek, Zachary Fisher

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 Parsing

The Coverage Principle: A Framework for Understanding Compositional Generalization

2025-05-26 · Hoyeon Chang, Jinho Park, Hanseul Cho, Sohee Yang 외

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…