NeuroLogic A*esque Decoding: Constrained Text Generation with Lookahead Heuristics
The dominant paradigm for neural text generation is left-to-right decoding from autoregressive language models. Constrained or controllable generation under complex lexical constraints, however, requires foresight to plan ahead feasible future paths. Drawing inspiration from the A* search algorithm, we propose NeuroLogic A*esque, a decoding algorithm that incorporates heuristic estimates of future cost. We develop efficient lookahead heuristics that are efficient for large-scale language models, making our method a drop-in replacement for common techniques such as beam search and top-k sampling. To enable constrained generation, we build on NeuroLogic decoding (Lu et al., 2021), combining its flexibility in incorporating logical constraints with A*esque estimates of future constraint satisfaction. Our approach outperforms competitive baselines on five generation tasks, and achieves new state-of-the-art performance on table-to-text generation, constrained machine translation, and keyword-constrained generation. The improvements are particularly notable on tasks that require complex constraint satisfaction or in few-shot or zero-shot settings. NeuroLogic A*esque illustrates the power of decoding for improving and enabling new capabilities of large-scale language models.
PDF Abstract NAACL 2022 PDF NAACL 2022 AbstractDatasets
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Text Generation | ROCStories | Beam search + A*esque (sample) | Perplexity | 2.16 | # 3 | |
BLEU-1 | 34.4 | # 1 | ||||
Text Generation | ROCStories | Beam search + A*esque (greedy) | Perplexity | 2.11 | # 1 | |
BLEU-1 | 34.3 | # 3 | ||||
Text Generation | ROCStories | Beam search | Perplexity | 2.24 | # 4 | |
BLEU-1 | 33.7 | # 4 | ||||
Text Generation | ROCStories | Beam search + A*esque (beam) | Perplexity | 2.14 | # 2 | |
BLEU-1 | 34.4 | # 1 |