What it takes to do recognizable research without having big names behind you (think unis or big research companies) and without prior track of publications?
If one wants to publish their results, how to make it to arXiv without endorsements?
Do someone has experience in being independent researcher?
You do not need to publish to journals/conferences if you are independent, and this is the most brilliant part! You can just post your results in your blog, or on a relevant forum, and people will cite you if those results are good. If it's in CS/ML/AI then posting a blog post on HN might reach more people and be recognized more than a preprint on arxiv.
For example, my post: https://k10v.github.io/2018/02/25/Solving-Bongard-problems-w... It's published only on a web page, and is cited by several papers and research groups, like https://arxiv.org/pdf/2206.07196.pdf
Recent quite impactful work in LLM interoperability by nostalgebraist is posted only on lesswrong, and now cited by many papers: https://www.lesswrong.com/posts/AcKRB8wDpdaN6v6ru/interpreti...
You can also publish any paper or witepaper to arXiv or openreview, you don't need any endorsements for that, but this is quite useless if you are not pursuing academic career.