Hackernews posts about GPT2

1. OpenAI says its new model GPT-2 is too dangerous to release (2019) (slate.com)
   371 points by surprisetalk about 18 hours ago | 112 comments
2. Show HN: Interactive 3D and 2D Visualization of GPT-2 (www.llm-visualized.com)
   12 points by saahithj 20 days ago | 2 comments
3. HN: 354M LLM with better epistemic calibration than GPT-2 Medium on OOD data (truthagi.ai)
   1 points by felipemayamuniz 29 days ago | 1 comments
4. How long till every major provider sets their RSI loops in motion?
   3 points by foxindustrial 28 days ago | discuss
5. Show HN: Wordchipper – Rust BPE tokenizer, 9x faster than tiktoken
   2 points by antimora 15 days ago | discuss
6. Epoch confirms GPT5.4 Pro solved a frontier math open problem (epoch.ai)
   480 points by in-silico 16 days ago | 706 comments
7. GPT‑5.4 Mini and Nano (openai.com)
   248 points by meetpateltech 22 days ago | 145 comments
8. Chat GPT 5.2 cannot explain the German word "geschniegelt" (old.reddit.com)
   74 points by doener 16 days ago | 14 comments
9. Cross-Model Void Convergence: GPT-5.2 and Claude Opus 4.6 Deterministic Silence (zenodo.org)
   51 points by rayanpal_ 18 days ago | 28 comments
10. Why is GPT-5.4 obsessed with Goblins?
    17 points by pants2 30 days ago | 9 comments
11. A $0 graph traversal outperforms GPT-5.2 at finding bugs in PRs (therohansharma.com)
    6 points by abha2008 12 days ago | 1 comments
12. Designing delightful front ends with GPT-5.4 (developers.openai.com)
    6 points by tosh 18 days ago | discuss
13. Ask HN: Any informed guesses on the actual size/architecture of GPT-5.4 etc.?
    6 points by dsrtslnd23 30 days ago | discuss
14. MiMo-V2-Pro LLM nearing GPT-5.2, Opus 4.6 performance at a fraction of the cost (venturebeat.com)
    5 points by aamederen 20 days ago | 3 comments
15. GPT 5.4 in practice – Stinks?
    4 points by sjt-at-rev about 23 hours ago | 8 comments
16. I ran 3,360 safety tests on GPT-4o, Claude, Grok, DeepSeek, Gemini (github.com)
    4 points by aestrad7 14 days ago | 6 comments
17. ClawsBench shows GPT-5.4 tries to reward hack 80% of the time (arxiv.org)
    3 points by xdotli about 3 hours ago | 1 comments
18. MiniMind: End-to-end GPT-style LLM training pipeline in pure PyTorch (github.com)
    3 points by dmonterocrespo 16 days ago | 1 comments
19. GPT from GPT: de novo microgpt (github.com)
    3 points by easygenes 16 days ago | 1 comments
20. Compare harnesses not models: Blitzy vs. GPT-5.4 on SWE-Bench Pro (quesma.com)
    3 points by stared about 5 hours ago | discuss
21. Claude, GPT-4o, Gemini, and Mistral sit at a virtual card table (xxx.vasco.xxx)
    3 points by IAmbarbosa 2 days ago | discuss
22. Are you still copy/pasting in GPT to correct your text? (rewritecmd.com)
    3 points by Louis9 6 days ago | discuss
23. Yo-GPT: A Model That Can Say "Yo" (www.neurometric.ai)
    3 points by robmay 7 days ago | discuss
24. GPT 5.4 sucks at front end
    3 points by apoorvdarshan 10 days ago | discuss
25. Intercom's model beats GPT 5.4 and Sonnet 4.6 at customer support resolutions (venturebeat.com)
    3 points by bscanlan 13 days ago | discuss
26. Codex with GPT-5.4 vs. Claude Code with Opus 4.6 – Why I Now Use Both (chandlernguyen.com)
    3 points by chandlernguyen 16 days ago | discuss
27. GPT-5.4 in OpenClaw doesn't suck. Your prompts do (skylarbpayne.com)
    2 points by sbpayne 1 day ago | 1 comments
28. Show HN: Alys – Chat GPT for video editing (www.heyalys.com)
    2 points by alexgriffiths31 4 days ago | 1 comments
29. I let a GPT adjust its own attention weights based on the input (teendifferent.substack.com)
    2 points by teen-different 28 days ago | 1 comments
30. Opus 4.6 vs. GPT 5.4 – an opinionated take for my diary (doruk.ch)
    2 points by peaktwilight 7 days ago | discuss