I read the paper. Seems like impressive stuff, with one major proviso: it doesn’t think up or do experiments. It looks at the literature and data sets, and tries to come up with a model for some phenomenon. If that model has features heretofore not noticed, great, maybe you’ve

[blogged] ResearchRabbit’s 2025 Revamp: iterative chaining without the clutter https://t.co/1HgRGROBO3 @RsrchRabbit

We just launched Walden, our rebuilt dataset built on a rewritten codebase. It's faster, cleaner, and adds 190M new works. Same schema, better data. Blog post: https://t.co/YEvsf8ToWW Full release notes:

Definitely picking up more from Research Commons. Too bad I think the profile currently isn't picking up automatically from Research commons matces (it can from core collection + preprint citation index). (3)

Join us Tonight for our FINALE!!!

Research commons + Preprint Citation index = Web of Science Platform coverage is much less selective. This is so though by default even if you search using "all databases" the default is to still exclude these two databases and you must explicitly remove the exclusion (2)

Very interesting web of science now as "Research Commons" Add 32M more metadata records, +21% more journal content! This pulls from open sources like @OpenAlex_org @CrossrefOrg https://t.co/Qnc2fvTxyw (1)

📢 Registration OPEN for #FORCE2026 (3–5 Jun 2026, Singapore 🇸🇬) Early-bird till 28 Feb 2026 Call for Proposals closes 9 Nov 2025 (AoE) Register: https://t.co/JUA9PLIDxX Submit proposal: https://t.co/N2fkFagFw5 #ScholarlyCommunication #ResearchCommunication

Realty Check

[Blogged] ] “We’re Good at Search”… Just Not the Kind That the AI era Demands - a Provocation

Another @NotebookLM ship - Huge quality improvements in our core Chat experience! When you need to work over very Long Context, Notebook is the best tool for your task. https://t.co/zHU1Ao6V0N

Things we learned @paperpile by asking 1,800 researchers about their use of AI tools in science: 1. ChatGPT is dominating and it's not even really close. Purpose-built apps for research are way down in the long tail. 1/n

Yup. There are many things wrong with ChatGPT. It's terrible for education, for starters. But the water thing is pure nonsense. I was glad to see Cal Newport say this recently too. He's a leading opponent of AI, but he's no fool.

Why do hundreds of people journey across the world to live and build together at Edge City? Week 1 of Edge City Patagonia, beautifully captured by Jeffrey Sun ☀️

🐇 The new ResearchRabbit is live! A big leap forward in how you explore and grow your research world 🌍✨ ✨ Better maps 🔍 Advanced search 🎨 Personalisation 📚 280M+ papers 💡 All designed to help you see further and stay curious 💜 👉 Explore: https://t.co/hA4pnHJ8wd 🔗

Final call for proposals for FORCE 2026 Scholarly Comm Conf in Singapore - June 2026. Proposals close 9 Nov. https://t.co/0MZYslDFzK - These are the current breakdowns of proposals by theme. Personally, I had hoped that would be more on "AI across research lifecycle".

✨ FORCE2026 is a leading conference on Scholarly communications → Extended deadline for proposals: November 9 → Special author rate for accepted proposals → Sponsorship opportunities still open https://t.co/kufefCs3fs #Force2026

The FORCE 2026 conference from 3-5 June 2026 will be hosted in Singapore from 3-5 June 2026. Call for proposals is open until 9 Nov - hope to see you in Singapore! (1)

How do bumble bees hums?

Read this. Mostly familiar with the arguments. The last few chapters drawing analogies with how hard things like launching space probes etc are, & saying while stopping AI development is hard, it is less so than fighting WW2 against axis powers -are new to me.

so maybe it doesn't have the capability or tool call to look at references.. who knows?

But is it "agentic enough" to do things like find me papers like X, but not referenced by X? I find ChatGPT can actually do that! But most of the academic agentic/deep search tools can't probably because it's agent workflow doesn't anticipate this request I think (2)

Yet another specialized academic search tool using the term "agentic". IMHO the issue is it doesn't really tell a user like me what it can actually do. For sure we know doing find me papers on topic x will work, but that would have worked prior to becoming "agent" (1)

Real prop traders need the best—and now one of the most trusted trading platforms for prop is about to get even better. Follow @NinjaTraderProp for updates.

This is a very interesting development , one can imagine the LLM looking at initial results (either from keyword or semantic) and "deciding" or give suggestions to adjust the keyword strategy? Essentially automating piloting the search strategy? Librarians will love this.