Hacker News

Article URL: https://www.reuters.com/world/pro-palestinian-student-permanent-us-resident-sues-halt-deportation-2025-03-24/

Comments URL: https://news.ycombinator.com/item?id=43470183

Points: 4

# Comments: 0

Article URL: https://www.ft.com/content/b4018fb0-8d52-46b4-984b-f776dfed3615

Comments URL: https://news.ycombinator.com/item?id=43470177

Points: 2

# Comments: 1

Article URL: https://github.com/habedi/hann

Comments URL: https://news.ycombinator.com/item?id=43470162

Points: 2

# Comments: 0

Article URL: https://www.nature.com/articles/d41586-025-00875-w

Comments URL: https://news.ycombinator.com/item?id=43469930

Points: 2

# Comments: 0

Article URL: https://news.northwestern.edu/stories/2025/03/new-carbon-negative-material-could-make-concrete-and-cement-more-sustainable/

Comments URL: https://news.ycombinator.com/item?id=43469928

Points: 2

# Comments: 0

Article URL: https://www.technologyreview.com/2025/03/25/1105488/how-to-delete-your-23andme-data/

Comments URL: https://news.ycombinator.com/item?id=43469915

Points: 2

# Comments: 0

Article URL: https://www.abc.net.au/news/science/2025-03-20/openai-generative-ai-hallucinations-chatbot-gpt45-test/105041122

Comments URL: https://news.ycombinator.com/item?id=43469910

Points: 2

# Comments: 0

Article URL: https://www.dabeaz.com/papers/CiSE/c5090.pdf

Comments URL: https://news.ycombinator.com/item?id=43469905

Points: 2

# Comments: 0

Article URL: https://matt.might.net/articles/phd-school-in-pictures/

Comments URL: https://news.ycombinator.com/item?id=43469898

Points: 2

# Comments: 0

Article URL: https://deadprogrammersociety.com/2025/03/calling-mcp-servers-the-hard-way.html

Comments URL: https://news.ycombinator.com/item?id=43469888

Points: 3

# Comments: 0

Article URL: https://www.lesswrong.com/posts/u8GMcpEN9Z6aQiCvp/rule-thinkers-in-not-out

Comments URL: https://news.ycombinator.com/item?id=43469885

Points: 2

# Comments: 0

Article URL: https://www.theguardian.com/artanddesign/2025/mar/11/zippos-circus-man-utd-manchester-united-norman-foster-new-stadium

Comments URL: https://news.ycombinator.com/item?id=43469878

Points: 1

# Comments: 0

Article URL: https://www.theguardian.com/technology/2025/mar/24/tesla-sales-eu-slump-elon-musks-donald-trump-byd

Comments URL: https://news.ycombinator.com/item?id=43469873

Points: 4

# Comments: 0

Article URL: https://matklad.github.io/2025/03/25/debugger-is-repl-is-debugger.html

Comments URL: https://news.ycombinator.com/item?id=43469866

Points: 1

# Comments: 0

Article URL: https://gist.github.com/haron/39852d7b05a924a64f620a7d189b3e86

Comments URL: https://news.ycombinator.com/item?id=43469856

Points: 1

# Comments: 0

We've built a simple SAML testing tool that lets you configure IDP and SP settings without any signup process. Perfect for quickly testing SAML implementations or experimenting with configurations. Key features: Configure IDP metadata, entity IDs, and redirect URLs Test SP settings (ACS URL, entity ID, attribute mappings) Optional SCIM configuration for directory syncing No accounts needed - just open and start testing Completely free to use Give it a try and let us know how we can improve it for the developer community!

Comments URL: https://news.ycombinator.com/item?id=43469843

Points: 1

# Comments: 0

Article URL: https://www.willbuckingham.com/plato-organist-to-the-beasts/

Comments URL: https://news.ycombinator.com/item?id=43469831

Points: 1

# Comments: 1

Article URL: https://huggingface.co/papers/2503.18102

Comments URL: https://news.ycombinator.com/item?id=43469821

Points: 2

# Comments: 0

Hey HN! I recently published an article titled "Spatial Text Rendering: Pushing the Limits of Spatial Understanding in LLMs" where I share a technique I've been using for quite some time now to help text-only LLMs process visually complex documents before Vision Language Models (VLMs) became usable. I thought it might be useful for anyone working with document processing!

Summary: This article introduces Spatial Text Rendering (STR), a method that bridges the gap between visually complex documents and text-only LLMs by preserving the crucial spatial information that gives documents their meaning. While Vision-Language Models (VLMs) continue to advance, we needed an immediate solution that could handle complex financial documents in the MEA region (but not limited to it), including Arabic text and mixed right-to-left scripts. STR uses image processing techniques to extract the document's underlying structure and render it as spatially-aware text that LLMs can understand.

Key Points and Highlights:

- Financial documents present unique challenges: complex layouts, mixed languages, and data that require absolute precision

- Spatial Text Rendering involves: document preprocessing/deskewing, OCR with spatial coordinates, structure extraction, and structural line detection

- We use a text-based rendering approach that translates visual structure into a format LLMs already understand from their pre-training

- A compaction process significantly reduces token usage while preserving key information

- Testing showed excellent results across multiple LLMs (Claude, GPT-4o, etc.) even without fine-tuning

- The approach offers an immediate solution for document processing while VLMs continue to develop and become more affordable to use

Side Open Discussion: One interesting aspect I've observed is that many LLMs seem to have robust spatial reasoning capabilities from their pre-training alone, despite not being explicitly trained for this task. This suggests that LLMs might have absorbed more spatial understanding through their text-only training than previously thought. I'm curious if others have observed and taken advantage of similar capabilities?

Let me know what you think!

Comments URL: https://news.ycombinator.com/item?id=43469804

Points: 1

# Comments: 0

Article URL: https://en.wikipedia.org/wiki/Cabbit

Comments URL: https://news.ycombinator.com/item?id=43469801

Points: 2

# Comments: 2