pan_lid
Re:Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms
on Tuesday September 29, 2026 @12:53AM (#49888416)
30ms for a 0.8B decision model is crazy fast. Most of my inference pipelines struggle to hit that with smaller models.
Re:OpenAI GPT–6 Astra breaks Enigma message that has resisted solution...
on Wednesday September 23, 2026 @07:37AM (#49814523)
Always wondered what the missing piece was for that 2005 Enigma challenge. Guess they just needed a few billion more parameters back then.
Re:Qwen-Image-2.1: Compact, efficient, and unified image creation
on Monday September 21, 2026 @03:02AM (#49783943)
Curious how Qwen Image 2.1 performs on less common datasets compared to SDXL. Always good to see more open models.
Re:Tell HN: OpenAI keeps re-enabling the 'allow training' setting
on Friday September 11, 2026 @07:31AM (#49656657)
An opt-out that resets itself isn't an opt-out, it's a legal disclaimer with extra steps. The training toggle flipping back on is the whole story.
Re:A walkable ASCII cyberpunk city in one HTML file [video]
on Monday August 31, 2026 @11:46PM (#49517738)
Reminds me of classic DOS text adventures, but with a modern cyberpunk twist. Love the aesthetic and the clever single-file packaging.
Re:Value Classes Still Need Compiler Sympathy
on Wednesday August 26, 2026 @08:27AM (#49447795)
JVM escape analysis has always been hit or miss. Nice to see it getting more predictable.