Paper ID: 2402.05785

Limits of Transformer Language Models on Learning to Compose Algorithms

Jonathan Thomm, Aleksandar Terzic, Giacomo Camposampiero, Michael Hersche, Bernhard Schölkopf, Abbas Rahimi

We analyze the capabilities of Transformer language models in learning compositional discrete tasks. To this end, we evaluate training LLaMA models and prompting GPT-4 and Gemini on four tasks demanding to learn a composition of several discrete sub-tasks. On both training LLaMA models from scratch and prompting on GPT-4 and Gemini, we measure how well these models can reuse primitives observable in the sub-tasks to learn the composition task. Our results indicate that compositional learning in state-of-the-art Transformer language models is highly sample inefficient: LLaMA requires more data samples than relearning all sub-tasks from scratch to learn the compositional task; in-context prompting with few samples is unreliable and fails at executing the sub-tasks or correcting the errors in multi-round code generation. Further, by leveraging complexity theory, we support these findings with a theoretical analysis focused on the sample inefficiency of gradient descent in memorizing feedforward models.

Submitted: Feb 8, 2024