Interlocking-free Selective Rationalization Through Genetic-based Learning
This addresses a key bottleneck in interpretable AI for researchers and practitioners by providing an interlocking-free solution, though it is incremental as it builds on existing select-then-predict pipelines.
The paper tackles the problem of interlocking in selective rationalization architectures, where one module dominates and leads to suboptimal performance, by introducing GenSPP, which uses genetic global search for disjoint training to avoid this issue. Experiments on synthetic and real-world benchmarks show that GenSPP outperforms state-of-the-art competitors.
A popular end-to-end architecture for selective rationalization is the select-then-predict pipeline, comprising a generator to extract highlights fed to a predictor. Such a cooperative system suffers from suboptimal equilibrium minima due to the dominance of one of the two modules, a phenomenon known as interlocking. While several contributions aimed at addressing interlocking, they only mitigate its effect, often by introducing feature-based heuristics, sampling, and ad-hoc regularizations. We present GenSPP, the first interlocking-free architecture for selective rationalization that does not require any learning overhead, as the above-mentioned. GenSPP avoids interlocking by performing disjoint training of the generator and predictor via genetic global search. Experiments on a synthetic and a real-world benchmark show that our model outperforms several state-of-the-art competitors.