Неофициальный Python-клиент для alphaxiv: как мы нашли скрытый API и упаковали его в пакет

У alphaxiv.org есть API — но найти его было непросто: публичная документация появилась совсем недавно, а до этого единственный способ разобраться в том, как он работает, — DevTools и живой трафик браузера. POST-запрос к api.alphaxiv.org/assistant/v2/chat , SSE-поток в ответе, модель aurelle-1 . На основе этого исследования мы собрали aurelle-py — Python-пакет для программного доступа к AI-ассистенту alphaxiv: задавать вопросы по arXiv-статьям, стримить ответы, встраивать в исследовательские пайплайны. Мы не первые, кто занялся этой темой, — но постарались сделать решение аккуратным и хорошо задокументированным. Что внутри: синхронный и асинхронный клиенты, SSE-парсер с независимым юнит-тестированием, Pydantic v2 для валидации, типизированные исключения ( AuthError , RateLimitError ), MCP-сервер для интеграции с Claude Desktop и Claude Code. pip install aurelle-py В статье — как мы нашли эндпоинт, разобрали формат запроса и ответа, какие ограничения выявили опытным путём и как устроен пакет внутри. GitHub: https://github.com/center4aai/aurelle-py

https://habr.com/ru/articles/1010046/

#llmагент #mcp #apiclient #arxiv #alphaxiv #claude_code

GitHub - center4aai/aurelle-py: Unofficial Python client for the alphaxiv.org Assistant API — AI-powered Q&A over arXiv papers

Unofficial Python client for the alphaxiv.org Assistant API — AI-powered Q&A over arXiv papers - center4aai/aurelle-py

GitHub

alphaXiv (@askalphaxiv)

NeurIPS에서 열린 Latent Space 팟캐스트에서 @swyx와 대화한 내용 요약. 발언자는 댓글 기반에서 심층 연구와 ML 샌드박스 환경으로의 진화, 그리고 alphaXiv를 AI 연구용 GitHub으로 만들겠다는 비전을 설명함. 팟캐스트 링크로 자세한 이야기와 향후 로드맵 확인 가능.

https://x.com/askalphaxiv/status/2016941098699280474

#podcast #neurips #alphaxiv #ai #research

alphaXiv (@askalphaxiv) on X

Really enjoyed chatting with @swyx ​on the @latentspacepod at NeurIPS! Walked through our origin story, evolution from comments towards deep research and ML sandbox environments, and vision for the future of alphaXiv as the GitHub for AI research 🚀 Check out the podcast below!

X (formerly Twitter)

#TIL that #AlphaXiv, has a collection of the datasets mentioned in the #AI papers in #ArXiv that is has indexed:

https://www.alphaxiv.org/?datasets=true

#AI #data

alphaXiv

Discuss, discover, and read arXiv papers.

alphaXiv

Bạn đang quan tâm đến nghiên cứu AI? AlphaXiv hoặcy cập tính năng notebookLM để xem bài커ng của arXiv! Chuyển đổi công thức phức tạp thành cuộc trò chuyện Génnerati Thật tuyệt.<br><br>#AI #AlphaXiv #arXiv #tool #nghldon

https://www.reddit.com/r/LocalLLaMA/comments/1oa0gid/alpharxiv/

On the Theoretical Limitations of Embedding-Based Retrieval | alphaXiv

View recent discussion. Abstract: Vector embeddings have been tasked with an ever-increasing set of retrieval tasks over the years, with a nascent rise in using them for reasoning, instruction-following, coding, and more. These new benchmarks push embeddings to work for any query and any notion of relevance that could be given. While prior works have pointed out theoretical limitations of vector embeddings, there is a common assumption that these difficulties are exclusively due to unrealistic queries, and those that are not can be overcome with better training data and larger models. In this work, we demonstrate that we may encounter these theoretical limitations in realistic settings with extremely simple queries. We connect known results in learning theory, showing that the number of top-k subsets of documents capable of being returned as the result of some query is limited by the dimension of the embedding. We empirically show that this holds true even if we restrict to k=2, and directly optimize on the test set with free parameterized embeddings. We then create a realistic dataset called LIMIT that stress tests models based on these theoretical results, and observe that even state-of-the-art models fail on this dataset despite the simple nature of the task. Our work shows the limits of embedding models under the existing single vector paradigm and calls for future research to develop methods that can resolve this fundamental limitation.

alphaXiv

alphaXiv is a very useful search engine for arXiv papers. One of its nice features is that for many papers, section headings are shown for the paper just below the abstract. If you click on the section heading, it will display some text from the paper relevant to that section of the paper.
It also usually shows a screenshot of the first page of the paper. It also shows discussion related to the paper.

https://www.alphaxiv.org/explore?custom-categories=information-extraction

#research #AI #ML #CS #alphaXiv

alphaXiv

Discuss, discover, and read arXiv papers.

alphaXiv
alphaXiv

Discuss, discover, and read arXiv papers.

✨ This movement towards a more transparent, inclusive, and collaborative scientific landscape is not just promising—it's transformative. It holds the potential to accelerate discovery, nurture innovation, and ensure that the benefits of research are universally accessible.

Together ✊

#OpenScience #ScientificInnovation #alphaXiv #Collaboration #Inclusivity #Transparency #PeerReview #GlobalResearch
#BioinformaticianNextDoor

Welcoming #alphaXiv to the Vanguard of Open Science and Transparent Peer Review!
In our era where collaboration, transparency, and inclusivity are not just values but necessities, we stand witness to the evolution in the landscape of peer review of scientific publications. #alphaXiv, is the latest archive joining the esteemed ranks of The PubPeer Foundation, Peer Community In In, and PREreview in championing the cause of open and transparent #peer_review.