SciRepEval: A Multi-Format Benchmark for Scientific Document Representations
This work addresses the problem of limited evaluation diversity for scientific document representations, benefiting researchers in natural language processing and information retrieval, though it is incremental as it builds on existing models like SPECTER and SciNCL.
The authors tackled the lack of diverse benchmarks for evaluating scientific document representations by introducing SciRepEval, a comprehensive benchmark with 24 tasks across four formats, and developed SPECTER2, a multi-format model that outperforms existing state-of-the-art methods by over 2 points absolute.
Learned representations of scientific documents can serve as valuable input features for downstream tasks without further fine-tuning. However, existing benchmarks for evaluating these representations fail to capture the diversity of relevant tasks. In response, we introduce SciRepEval, the first comprehensive benchmark for training and evaluating scientific document representations. It includes 24 challenging and realistic tasks, 8 of which are new, across four formats: classification, regression, ranking and search. We then use this benchmark to study and improve the generalization ability of scientific document representation models. We show how state-of-the-art models like SPECTER and SciNCL struggle to generalize across the task formats, and that simple multi-task training fails to improve them. However, a new approach that learns multiple embeddings per document, each tailored to a different format, can improve performance. We experiment with task-format-specific control codes and adapters and find they outperform the existing single-embedding state-of-the-art by over 2 points absolute. We release the resulting family of multi-format models, called SPECTER2, for the community to use and build on.