Perplexity
perplexity.ai
Answer engine that links every claim to a source. Handy for finding candidate URLs to add to a notebook — import the cited pages, not the AI summary itself.
Perplexity answers questions by pulling from live web sources and citing each one, a fast way to generate a list of candidate sources on an unfamiliar topic rather than working through a single search results page by hand. For a notebook, the value is in the cited pages it points to — import those, not Perplexity's own summary, since the summary is a synthesis you can't independently verify the way you can a primary source. Its citations skew toward whatever ranks well and is freely accessible, so it can miss paywalled or more obscure academic sources a targeted search wouldn't.
Import into Gemini Notebook
- 1Copy the link above (or a specific page from the site).
- 2In your notebook, click Add source → Website.
- 3Paste the link and confirm — several links can be pasted at once, one per line.
Conference proceedings and lab blogs publish full papers as PDFs or Website pages — those import cleanly and keep citations intact.
More ai & technology research sources
- AAAI Conference on Artificial IntelligenceProceedings of a major annual AI conference covering the whole field. Individual paper PDFs from the proceedings work well as sources.
- ACM Digital LibraryPublisher library for computing research — conference proceedings and journals. Many papers are open access, and those PDFs import directly.
- AI Alignment ForumLong-form technical posts on AI safety and alignment research. Individual post URLs work well as Website sources.
- DistillInteractive, peer-reviewed explainers of machine learning concepts. Dormant since 2021, but the articles import well as Website sources for ML study notebooks.
- Google DeepMind ResearchPublications and technical blog posts from Google DeepMind. Blog write-ups often summarize a paper more digestibly than the PDF — add both as sources.
- ICLR (OpenReview)ICLR papers with their full peer-review threads on OpenReview. Reviews and author rebuttals add useful critical context alongside the paper PDF.
- IEEE XploreJournals, conference proceedings, and standards in electrical engineering and computing. Abstracts are open; full PDFs need a subscription or purchase.
- Journal of Machine Learning Research (JMLR)Long-running open-access machine learning journal. Every paper is a free PDF with no paywall, making it easy to build ML literature notebooks.
- Linguistic Data ConsortiumDistributes speech and text corpora for language research and model training. Access is membership-based; corpus documentation pages are freely viewable.
- MLCommons / MLPerf BenchmarksHome of the MLPerf benchmarks for training and inference speed. Results pages and methodology docs anchor notebooks comparing AI hardware and systems.
- Nature Machine IntelligencePeer-reviewed journal on machine learning and AI applications. A curated counterweight to preprints when building an AI literature notebook.
- NeurIPS ProceedingsEvery paper from the top machine learning conference, back to 1987. Free PDFs make it easy to build a notebook around a research thread across years.
- OpenMLShared machine learning datasets and benchmark task results with API access. Dataset description pages document provenance for reproducibility notebooks.
- OpenReview.netPapers plus their full peer-review threads for venues like ICLR. Importing reviews alongside the paper gives a notebook both sides of the argument.
- Proceedings of Machine Learning Research (PMLR)Free full-text papers from ICML, AISTATS, UAI and other ML venues. Direct PDF links with no paywall — easy bulk imports for ML literature reviews.
- The GradientLong-form essays on AI research written for a technical but general audience. Articles pair well with the papers they discuss in the same notebook.
- Transactions on Machine Learning Research (TMLR)Open-access ML journal with rolling submissions and public reviews. Each paper has a landing page plus a PDF you can pull straight into a notebook.