On the Zero-Shot Generalization of Machine-Generated Text Detectors
This addresses the challenge of detecting machine-generated text from unseen generators, which is crucial for combating misinformation and ensuring content authenticity, though it is incremental in improving detector robustness.
The study investigated the zero-shot generalization of machine-generated text detectors across different large language models, finding that detectors trained on medium-sized models could generalize to larger versions, enabling robust detection through ensemble methods.
The rampant proliferation of large language models, fluent enough to generate text indistinguishable from human-written language, gives unprecedented importance to the detection of machine-generated text. This work is motivated by an important research question: How will the detectors of machine-generated text perform on outputs of a new generator, that the detectors were not trained on? We begin by collecting generation data from a wide range of LLMs, and train neural detectors on data from each generator and test its performance on held-out generators. While none of the detectors can generalize to all generators, we observe a consistent and interesting pattern that the detectors trained on data from a medium-size LLM can zero-shot generalize to the larger version. As a concrete application, we demonstrate that robust detectors can be built on an ensemble of training data from medium-sized models.