Transformer-based masked language models such as BERT, trained on general corpora, have shown impressive performance on downstream tasks. It has also been demonstrated that the downstream task performance of such models can be improved by pretraining larger models for longer on more data. In this work, we empirically evaluate the extent to which these results extend to tasks in science. We use 14 domain-specific transformer-based models (including ScholarBERT, a new 770Mparameter science-focused masked language model pretrained on up to 225B tokens) to evaluate the impact of training data, model size, pretraining and finetuning time on 12 downstream scientific tasks. Interestingly, we find that increasing model size, training data, or compute time does not always lead to significant improvements (i.e., >1% F1), if any, in scientific information extraction tasks. We offer possible explanations for this surprising result.

Answering multi-hop reasoning questions requires retrieving and synthesizing information from diverse sources. Large Language Models (LLMs) struggle to perform such reasoning consistently. Here we propose an approach to pinpoint and rectify multi-hop reasoning failures through targeted memory injections on LLM attention heads. First, we analyze the per-layer activations of GPT-2 models in response to single and multi-hop prompts. We then propose a mechanism that allows users to inject pertinent prompt-specific information, which we refer to as “memories,” at critical LLM locations during inference. By thus enabling the LLM to incorporate additional relevant information during inference, we enhance the quality of multi-hop prompt completions. We show empirically that a simple, efficient, and targeted memory injection into a key attention layer can often increase the probability of the desired next token in multi-hop tasks, by up to 424%.