| 1. | | Are two heads better than one? (eieio.games) |
| 210 points by evakhoury 5 days ago | 72 comments |
|
| 2. | | Why some clothes shrink in the wash and how to unshrink them (swinburne.edu.au) |
| 534 points by OptionOfT 8 days ago | 279 comments |
|
| 3. | | Everything you never wanted to know about file locking (2010) (apenwarr.ca) |
| 90 points by SmartHypercube 10 days ago | 17 comments |
|
| 4. | | Reproducing DeepSeek's MHC: When Residual Connections Explode (taylorkolasinski.com) |
| 121 points by taykolasinski 6 days ago | 30 comments |
|
| 5. | | You are not required to close your <p>, <li>, <img>, or <br> tags in HTML (novalistic.com) |
| 170 points by jen729w 9 days ago | 228 comments |
|
| 6. | | Sampling at negative temperature (cavendishlabs.org) |
| 203 points by ag8 7 days ago | 60 comments |
|
| 7. | | Meditation as Wakeful Relaxation: Unclenching Smooth Muscle (psychotechnology.substack.com) |
| 192 points by surprisetalk 11 days ago | 141 comments |
|
| 8. | | How to code Claude Code in 200 lines of code (mihaileric.com) |
| 814 points by nutellalover 10 days ago | 238 comments |
|
| 9. | | The Q, K, V Matrices (arpitbhayani.me) |
| 226 points by yashsngh 11 days ago | 79 comments |
|
| 10. | | JavaScript's For-Of Loops Are Fast (waspdev.com) |
| 70 points by surprisetalk 15 days ago | 71 comments |
|
| 11. | | A Technical Tour of the DeepSeek Models from V3 to v3.2 (sebastianraschka.com) |
| 8 points by giuliomagnifico 46 days ago |
|
| 12. | | Why does a least squares fit appear to have a bias when applied to simple data? (stats.stackexchange.com) |
| 306 points by azeemba 14 days ago | 75 comments |
|
| 13. | | FracturedJson (github.com/j-brooke) |
| 575 points by PretzelFisch 16 days ago | 164 comments |
|
| 14. | | Animated AI (animatedai.github.io) |
| 302 points by frozenseven 23 days ago | 28 comments |
|
| 15. | | What an unprocessed photo looks like (maurycyz.com) |
| 2510 points by zdw 21 days ago | 409 comments |
|
| 16. | | Openhaystack: Build 'AirTags' – track Bluetooth devices via Apple's network (github.com/seemoo-lab) |
| 459 points by thunderbong 11 months ago | 113 comments |
|
| 17. | | Three ways to solve problems (andreasfragner.com) |
| 145 points by 42point2 28 days ago | 22 comments |
|
| 18. | | Where Is GPT in the Chomsky Hierarchy? (fi-le.net) |
| 55 points by fi-le 35 days ago | 76 comments |
|
| 19. | | I've been writing ring buffers wrong all these years (2016) (snellman.net) |
| 165 points by flaghacker 33 days ago | 67 comments |
|
| 20. | | If a Meta AI model can read a brain-wide signal, why wouldn't the brain? (1393.xyz) |
| 141 points by rdgthree 36 days ago | 92 comments |
|
| 21. | | How to build a IP geolocation database from scratch? (ipapi.is) |
| 458 points by incolumitas on Sept 14, 2023 | 173 comments |
|
| 22. | | Touching the Elephant – TPUs (considerthebulldog.com) |
| 199 points by giuliomagnifico 43 days ago | 63 comments |
|
| 23. | | The inefficiency of RL, and implications for RLVR progress (dwarkesh.com) |
| 120 points by cubefox 52 days ago | 48 comments |
|
| 24. | | Agent design is still hard (pocoo.org) |
| 426 points by the_mitsuhiko 57 days ago | 258 comments |
|
| 25. | | Ask HN: How are Markov chains so different from tiny LLMs? |
| 204 points by JPLeRouzic 62 days ago | 189 comments |
|
| 26. | | How do the pros get someone to leave a cult? (theguardian.com) |
| 139 points by n1b0m 60 days ago | 124 comments |
|
| 27. | | Ditch your mutex, you deserve better (chrispenner.ca) |
| 139 points by commandersaki 68 days ago | 203 comments |
|
| 28. | | Solving a million-step LLM task with zero errors (arxiv.org) |
| 222 points by Anon84 61 days ago | 95 comments |
|
| 29. | | Refusal in LLMs is mediated by a single direction (lesswrong.com) |
| 110 points by veryluckyxyz on May 3, 2024 | 20 comments |
|
| 30. | | Heretic: Automatic censorship removal for language models (github.com/p-e-w) |
| 745 points by melded 63 days ago | 380 comments |
|
|
| More |
