I have never been a fan of Google Scholar, the tool designed for searching scientific publications. It is said that the main advantage of this platform is its vast database of scientific articles. However, I believe that is precisely the problem. I find the interface terrible and the filtering options inadequate, making it difficult to extract relevant results from the overwhelming volume of publications across various fields.
The Basics
Personally, when I need to find specific information, my go-to search database is PubMed. I also sometimes use the French platform HAL, which includes a collection of dissertations defended in France.
For keeping up with the latest research, I keep an eye on preprint platforms like bioRxiv, psyArXiv, and medRxiv. In fact, I recently learned about a new database dedicated to retracted preprints from these arXiv servers: WithdrarXiv, a resource aimed at providing insights and statistics on the reasons behind preprint retractions.
A relevant article in Nature:
Rao, D., Young, J., Dietterich, T., & Callison-Burch, C. (2024). WithdrarXiv: A Large-Scale Dataset for Retraction Study (arXiv:2412.03775). arXiv. https://doi.org/10.48550/arXiv.2412.03775
The WithdrarXiv database itself (which is also a preprint):
Singh Chawla, D. (2025). ‘WithdrarXiv’ database of 14,000 retracted preprints launches. Nature. https://doi.org/10.1038/d41586-025-00011-8
AI-Based Tools
For Google Scholar’s 20th anniversary, Nature published a general-audience article on… AI-powered tools that could potentially replace Google Scholar. "If there was ever a moment when Google Scholar could be dethroned as the leading academic search engine, it might be now," says Jevin West, an expert in computational social sciences.
Mallapaty, S. (2024). Can Google Scholar survive the AI revolution? Nature, 635(8040), 797–798. https://doi.org/10.1038/d41586-024-03746-y
This is a great opportunity to update the list of tools you and I can use for literature research! I have been using specialized tools based on large language models (LLMs) for some time now, such as Elicit, Scite.ai (back when it was free…), and Versa, an internal UCSF tool powered by ChatGPT, designed to be secure and trained on scientific literature.
However, through this article, I discovered two more AI-driven research tools:
Consensus, an AI-based search engine
Undermind, which functions more like a research assistant
I plan to test them in the coming days!
An LLM Specifically for Neuroscience
Finally, here’s a little bonus for neuroscience researchers: BrainGPT.org! This is a generative model that takes things a step further by synthesizing data from neuroscience literature. It aims to suggest study designs and generate likely data patterns, making it easier to evaluate models and identify unexpected results. I don’t know much more about it yet, but I signed up to test it and can’t wait to try it out!
Luo, X., Rechardt, A., Sun, G., Nejad, K. K., Yáñez, F., Yilmaz, B., Lee, K., Cohen, A. O., Borghesani, V., Pashkov, A., Marinazzo, D., Nicholas, J., Salatiello, A., Sucholutsky, I., Minervini, P., Razavi, S., Rocca, R., Yusifov, E., Okalova, T., … Love, B. C. (2024). Large language models surpass human experts in predicting neuroscience results. Nature Human Behaviour, 1–11. https://doi.org/10.1038/s41562-024-02046-9
This article was initially published in French, and was automatically translated using ChatGPT.
What About You?
What are your favorite research tools? Do you have any other recommendations? Share them in the comments—I’d love to discover your suggestions!
Comments