HN Explorer

rochansinha

↗ HN profile [ 115 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Yes we use feedback and de-ID'ed data to improve ChatGPT– like every LLM company (xcancel.com)
    1 points 0
  2. Nvidia AVO achieves 100% in ARC-AGI-3 (developer.nvidia.com)
    3 points 2
  3. DeepSeek-V4-Flash-Vision-Exp Is Now Live on the DeepSeek API Platform (twitter.com)
    1 points 1
  4. Arc AGI-3 100% solved with a skill (twitter.com)
    1 points 0
  5. Analog Cognition and Consciousness (doi.org)
    1 points 1
  6. DeepSeek Harness Released (twitter.com)
    10 points 0
  7. 1 points 0
  8. ZoomIt for Mac (github.com)
    3 points 0
  9. A puzzle to test your abstract reasoning skills (decryptiq.net)
    2 points 2
  10. Agent Plugins (agent-plugins.org)
    3 points 0
  11. Open AI announces two new models- GPT-transcribe and GPT-live-transcribe (developers.openai.com)
    3 points 1
  12. DOE announces first Genesis Mission projects for AI-driven scientific discovery (energy.gov)
    5 points 1
  13. The hard part begins after the lab experiments end (rochansinha.com)
    2 points 0
  14. Why China got rich and India didn't (davidoks.blog)
    82 points 35
  15. How do you use LLMs to verify databases with minimal hallucinations?

    I have a medium sized database of around 150 entries each with 10-15 parameters. It was put together by Claude. But the amount of hallucinated data is extraordinary! Trying to fix it using another llm like chat gpt or…

    1 points 1
  16. Why Water Is the Real Achilles Heel of the Chip Market (macronotes.substack.com)
    6 points 0
  17. 22 points 0
  18. The Future of Veritasium (youtube.com)
    2 points 1
  19. Skills Officially Comes to Codex (developers.openai.com)
    304 points 34