The Turking Test: Can Language Models Understand Instructions?
This addresses the problem of evaluating instruction understanding in AI, which could offer an alternative to few-shot learning, but the results are incremental as they highlight current limitations without a breakthrough solution.
The paper introduces the Turking Test to assess if language models can follow natural language instructions of varying complexity, from simple tasks to creative ones, and finds that a large pretrained model performs poorly across all tasks, often ignoring instructions.
Supervised machine learning provides the learner with a set of input-output examples of the target task. Humans, however, can also learn to perform new tasks from instructions in natural language. Can machines learn to understand instructions as well? We present the Turking Test, which examines a model's ability to follow natural language instructions of varying complexity. These range from simple tasks, like retrieving the nth word of a sentence, to ones that require creativity, such as generating examples for SNLI and SQuAD in place of human intelligence workers ("turkers"). Despite our lenient evaluation methodology, we observe that a large pretrained language model performs poorly across all tasks. Analyzing the model's error patterns reveals that the model tends to ignore explicit instructions and often generates outputs that cannot be construed as an attempt to solve the task. While it is not yet clear whether instruction understanding can be captured by traditional language models, the sheer expressivity of instruction understanding makes it an appealing alternative to the rising few-shot inference paradigm.