Elicit was the first AI research tool to analyze data in tables, which we launched two years ago. We’re excited to now be the first to systematically apply frontier multimodal reasoning to help scientists. Many researchers review the figures first before diving deeper into the paper. You can now do that more seamlessly in Elicit’s paper reader. 
When you create a systematic review, Elicit will extract data from figures of up to 1,000 papers. 
When you run an Elicit Report, you can see which figures are used to support claims in the report. 
Figures contain critical information not always described in the text of papers. Many researchers can determine the quality or relevance of a paper just by skimming the figures. Elicit now understands figures!
Elicit is the first AI tool that can systematically parse, interpret, and extract data from figures across thousands of papers. That includes Kaplan-Meier curves, heatmaps, reaction schemes, and microscopy images. 
css/html nerds, is there a reason why code highlighting implementations haven't switched to the css custom highlight api yet? browser support? nobody has implemented it yet?
on the surface, seems like it'd be a huge boon for smaller dom trees developer.mozilla.org/en-US/docs/W... how to get cancelled over this tweet: - post it - someone qt'ed, taking it very seriously and saying it was bad advice - i self-replied to clarify that it was a joke - self-reply was considered mansplaining (?) - dogpile re: mansplaining (?)
thinking back to the time i almost got cancelled over this tweet
Today @rabernat.bsky.social is at the @nationalacademies.org in DC participating on a panel about governance models for Earth data with @ecmwf.int, @esipfed.bsky.social, @wmo-global.bsky.social and more: we only wish we all could attend "Future Directions for Earth Observations and Stewardship" 
Berliners, Claude Code Anonymous is back for round 2 on Wednesday 10 December! Come and trade agentic-coding stories about wins, failures, and lessons learned, joined by @steipete.me 😎
The call for lightning talks is open too. Apply when you register 👇
luma.com/7xp4jpqh 
The only main difference is default print width. Other than that it should not produce massive diffs.
You should! At the bottom. The new default is a height of 40 lines
great! hope you like it, lmk if you have any other questions :))
That’s cool, happy to help! An individual shell session, so from when that specific shell started up - usually a terminal tab, split, etc We basically just generate a new UUID for each session in your .zshrc
the self update is actually a separate binary called atuin-update, that's installed by our install script. it can also be called as `atuin update`. even then, it won't self update automatically if you don't have that binary, then it can't self update + will rely on the package manager.
It's in most system package managers (homebrew, apt, etc), but can self update if you use our install script. We had a _lot_ of users who didn't know how to use their package manager, and a lot who complained that `atuin update` didn't work
1. yep that's right 2. it toggles filter mode - these filter search by global, host, current directory or current session. you can press ctrl-r to cycle through these 3. we mostly have this, but call it command chaining. It'll automatically append if the end of the current line is && or ||
ahhh right! yeah we have that now, and a few other things. it's the default - enter to execute, tab to insert it into the line editor
awh no :/ we've made some tweaks that improve muscle memory breakage, depending on when you tried it! was it anything in particular?
🫶🫶🫶
@atuin.sh does what you're looking for! lmk if you have any questions 
PS: Oxfmt will power the `vite fmt` command in Vite+ 👀
This first alpha release focuses on JavaScript and TypeScript, with support for additional languages coming soon. Also expect built-in sorting and more configuration options in the future.