Differentiable Tree Operations Promote Compositional Generalization
This addresses the problem of compositional generalization in AI for tasks like semantic parsing and language generation, representing a novel method rather than an incremental improvement.
The paper tackled the challenge of learning sequences of discrete symbolic operations for structure-to-structure transformation tasks by introducing a differentiable tree interpreter and a Differentiable Tree Machine (DTM) architecture, achieving 100% out-of-distribution compositional generalization on synthetic tasks compared to baselines scoring less than 30%.
In the context of structure-to-structure transformation tasks, learning sequences of discrete symbolic operations poses significant challenges due to their non-differentiability. To facilitate the learning of these symbolic sequences, we introduce a differentiable tree interpreter that compiles high-level symbolic tree operations into subsymbolic matrix operations on tensors. We present a novel Differentiable Tree Machine (DTM) architecture that integrates our interpreter with an external memory and an agent that learns to sequentially select tree operations to execute the target transformation in an end-to-end manner. With respect to out-of-distribution compositional generalization on synthetic semantic parsing and language generation tasks, DTM achieves 100% while existing baselines such as Transformer, Tree Transformer, LSTM, and Tree2Tree LSTM achieve less than 30%. DTM remains highly interpretable in addition to its perfect performance.