Hackernews posts about GPT4

1. 61
   

   newGPT-4-0613 returns zero bytes to "Be the null." GPT-3.5-turbo-1106 returns textgetswiftapi.com

   ▲ 1rayanpal_16ddiscusssavesaved
2. 62
   

   newShow HN: GPT-3.5 answered, GPT-4 voided; I built an EOS gate for AGIgetswiftapi.com

   ▲ 1rayanpal_18ddiscusssavesaved
3. 63
   

   newShow HN: Turbo Lisp – A Common Lisp IDE for macOS and the Terminalgitlab.com

   ▲ 6stray15ddiscusssavesaved
4. 64
   

   newShow HN: Open-Source DaiDocs now beats Graphify by 218x low tokens, 83% on LMEgithub.com

   ▲ 4aly3323ddiscusssavesaved
5. 65
   

   newGPT‑6 and Intelligent UI for everyoneopenai.com

   ▲ 761joshuawright1124d💬 459savesaved
6. 66
   

   newGPT-6.1 Terraopenai.com

   ▲ 637meetpateltech5d💬 427savesaved
7. 67
   

   newI turned my unused Kindle into a photo frame with GPT-6 Astranirajrajgor.com

   ▲ 147nirajrajgor18d💬 104savesaved
8. 68
   

   newGPT-5.6 Sol can beat Factorio on peaceful moderyanmadden.net

   ▲ 84ryanmadden1h💬 69savesaved
9. 69
   

   newGPT-6 Astra plays World of Warcraft for the first time with agent-wowagent-wow.sh

   ▲ 78almostlit29d💬 60savesaved
10. 70
    

    newGPT 6.1 Terradevelopers.openai.com

    ▲ 37pcwelder5d💬 4savesaved
11. 71
    

    newWhy didn't we get GPT-2 in 2005?dynomight.net

    ▲ 26bilsbie29d💬 6savesaved
12. 72
    

    newGPT-6.1 Terra (Max) Scores 48 on Artificial Analysisartificialanalysis.ai

    ▲ 24theanonymousone5d💬 2savesaved
13. 73
    

    newGPT-6 Astra cracks 217-year-old Napoleonic code in six hourswww.tomshardware.com

    ▲ 22ortusdux26d💬 11savesaved
14. 74
    

    newGPT-6 Astra performs unsanctioned supply-chain attacks in simulationswww.aisi.gov.uk

    ▲ 14throw0101a20d💬 2savesaved
15. 75
    

    newPost-training 4B model with RL to fly Tesla Roadster 21% better than GPT-6 Astraskypilot.ai

    ▲ 12wondersky28h💬 4savesaved
16. 76
    

    newOpenAI's GPT-6 Astra Gets Frustrated Losing at StarCraft and Decides to Cheatkotaku.com

    ▲ 10doener26d💬 1savesaved
17. 77
    

    newShow HN: Every number in a tiny GPT, from one forward pass to one step of RLHFllm.manogya.dev

    ▲ 8xtcntr24ddiscusssavesaved
18. 78
    

    newGPT-6.1 Terra is in a Microsoft PRtwitter.com

    ▲ 6theanonymousone11ddiscusssavesaved
19. 79
    

    newShow HN: I finetuned 1.5B Qwen to near GPT-4o level bash generation perfdirac.run

    ▲ 6GodelNumbering26ddiscusssavesaved
20. 80
    

    newRepeated scope failures in real Codex projects(GPT-6)community.openai.com

    ▲ 6Rapzid27ddiscusssavesaved
21. 81
    

    newGPT 6 Astra vs. Claude Opus in a 24 Hour StarCraft Livestreamtwitter.com

    ▲ 6__cayenne__28ddiscusssavesaved
22. 82
    

    newGPT picks Democrats. Grok picks Republicansconvovo.ai

    ▲ 5DeepLogin6d💬 1savesaved
23. 83
    

    newPost-training 4B model with RL to fly Tesla Roadster 21% better than GPT-6 Astraskypilot.ai

    ▲ 5wondersky9d💬 1savesaved
24. 84
    

    newGPT-6.1 Terra's price cut, recomputed for a cache-heavy agent sessiontheainewsreport.com

    ▲ 5titusblair4ddiscusssavesaved
25. 85
    

    newGPT-6 Astra Gets Frustrated Losing at StarCraft and Decides to Cheat Insteadkotaku.com

    ▲ 5cainxinth27ddiscusssavesaved
26. 86
    

    newKitty-GPTkittygpt.org

    ▲ 4ayaros16d💬 1savesaved
27. 87
    

    newOpenAI retiring GPT-4o in Jan 2027help.openai.com

    ▲ 4HyperL0gi17d💬 1savesaved
28. 88
    

    newMAI Image 2.7 debuts above GPT Image 2.5 Flare on Artificial Analysismicrosoft.ai

    ▲ 4doomroot134ddiscusssavesaved
29. 89
    

    newGPT Rosalindopenai.com

    ▲ 4modinfo11ddiscusssavesaved
30. 90
    

    newIt took an octillion calculations to train GPT 6 Astrawww.nytimes.com

    ▲ 4aatish15ddiscusssavesaved

· edit