| 1. | | Hypura – A storage-tier-aware LLM inference scheduler for Apple Silicon (github.com/t8) |
| 221 points by tatef 27 days ago | 86 comments |
|
| 2. | | Show HN: Duplicate 3 layers in a 24B LLM, logical deduction .22→.76. No training (github.com/alainnothere) |
| 265 points by xlayn 33 days ago | 80 comments |
|
| 3. | | Executing programs inside transformers with exponentially faster inference (percepta.ai) |
| 330 points by u1hcw9nx 39 days ago | 130 comments |
|
| 4. | | When AI writes the software, who verifies it? (leodemoura.github.io) |
| 305 points by todsacerdoti 48 days ago | 298 comments |
|
| 5. | | MCP server that reduces Claude Code context consumption by 98% (mksg.lu) |
| 570 points by mksglu 51 days ago | 107 comments |
|
| 6. | | Agent orchestration for the timid (substack.com) |
| 128 points by markferree 86 days ago | 32 comments |
|
| 7. | | LED lighting undermines visual performance unless supplemented by wider spectra (nature.com) |
| 212 points by bookofjoe 85 days ago | 211 comments |
|
| 8. | | Ask HN: How are you doing RAG locally? |
| 413 points by tmaly 3 months ago | 156 comments |
|
| 9. | | Lessons from 14 years at Google (addyosmani.com) |
| 1678 points by cdrnsf 3 months ago | 688 comments |
|
| 10. | | The Burrows-Wheeler Transform (sandbox.bio) |
| 157 points by g0xA52A2A 6 months ago | 40 comments |
|
| 11. | | Getting AI to work in complex codebases (github.com/humanlayer) |
| 517 points by dhorthy 6 months ago | 418 comments |
|
| 12. | | LLM Visualization (bbycroft.net) |
| 640 points by gmays 7 months ago | 46 comments |
|
| 13. | | Vector search on our codebase transformed our SDLC automation (medium.com/antonybrahin) |
| 34 points by antonybrahin 7 months ago | 15 comments |
|
| 14. | | Understanding Transformers Using a Minimal Example (rti.github.io) |
| 295 points by rttti 7 months ago | 25 comments |
|
| 15. | | Show HN: Fine-tuned Llama 3.2 3B to match 70B models for local transcripts (bilawal.net) |
| 31 points by phantompeace 7 months ago | 8 comments |
|
| 16. | | 6 weeks of Claude Code (puzzmo.com) |
| 581 points by mike1o1 8 months ago | 590 comments |
|
| 17. | | Andrew Ng: Building Faster with AI [video] (youtube.com) |
| 302 points by sandslash 9 months ago | 96 comments |
|
| 18. | | The time bomb in the tax code that's fueling mass tech layoffs (qz.com) |
| 1575 points by booleanbetrayal 10 months ago | 948 comments |
|
| 19. | | InventWood is about to mass-produce wood that's stronger than steel (techcrunch.com) |
| 487 points by LorenDB 11 months ago | 432 comments |
|
| 20. | | Run LLMs on Apple Neural Engine (ANE) (github.com/anemll) |
| 286 points by behnamoh 11 months ago | 123 comments |
|
| 21. | | Grappling with Infinity in Constraint Solvers (tuzz.tech) |
| 26 points by ColinWright on April 9, 2025 |
|
| 22. | | Supervisors often prefer rule breakers, up to a point (aom.org) |
| 198 points by rustoo on April 2, 2025 | 154 comments |
|
| 23. | | The long-awaited Friend Compound laws in California (supernuclear.substack.com) |
| 156 points by simonebrunozzi on March 26, 2025 | 210 comments |
|
| 24. | | Pen and Paper Exercises in Machine Learning (2022) (arxiv.org) |
| 413 points by ibobev on March 21, 2025 | 58 comments |
|
| 25. | | The Defer Technical Specification: It Is Time (thephd.dev) |
| 131 points by mattjhall on March 16, 2025 | 68 comments |
|
| 26. | | Ladder: Self-improving LLMs through recursive problem decomposition (arxiv.org) |
| 370 points by fofoz on March 7, 2025 | 110 comments |
|
| 27. | | Ask HN: Do your eyes bug you even though your prescription is "correct"? |
| 340 points by jbornhorst on March 7, 2025 | 326 comments |
|
| 28. | | Wi-Fi 7 Overview (2024) (mrncciew.com) |
| 44 points by speckx on Feb 26, 2025 | 46 comments |
|
| 29. | | I built an open source AI tool to find my autoimmune disease (reddit.com) |
| 412 points by makehistory on Feb 10, 2025 | 300 comments |
|
| 30. | | Show HN: A website that heatmaps your city based on your housing preferences (theretowhere.com) |
| 321 points by WiggleGuy on Feb 7, 2025 | 99 comments |
|
|
| More |
