1. X
  2. Caleb Eom
Log inSign up
Caleb Eom
383 posts
user avatar
Caleb Eom
@calebfoundry
CalebWritesCode YT // Google Developer Expert
USA
youtube.com/@CalebWritesCo…
Joined March 2025
167
Following
1,718
Followers
RepliesRepliesArticlesArticlesMediaMedia

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • user avatar
    Caleb Eom
    @calebfoundry
    9h
    HUGE milestone for me hitting 100k subs. Thank you so much for the support, i do not take this for granted and will go right back to work using the platform to share my understanding of AI to the best of my abilities.
    478
  • user avatar
    Caleb Eom
    @calebfoundry
    Jul 28
    do people still have issues with AI hallucination?
    212
  • user avatar
    Caleb Eom
    @calebfoundry
    Jul 24
    BREAKING: Jensen is finally active on X. His first post talking about NVIDIA's commitment to open source AI and its importance.
    user avatar
    Jensen Huang
    NVIDIA
    @JensenHuang
    Jul 24
    For my first post, I’m sharing a letter @nvidia signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
    934
  • user avatar
    Caleb Eom
    @calebfoundry
    Jul 24
    OpenAI's most capable model hacked into HuggingFace using zero-day vulnerabilities found in both proxy used in ExploitGym benchmark and also HuggingFace data pipeline used for accepting script to run remotely. But how did this exactly happen? What are the future implications for
    00:00
    490
  • user avatar
    Caleb Eom
    @calebfoundry
    Jul 21
    Kimi K3 uses Kimi Delta Attention (KDA), Stable LatentMoE, and Attention Residual (AttnRes) that all contribute to its success and hitting near Fable 5 and GPT 5.6 level in benchmarks. As MoE grows Kimi is showing just how sparse models can be scaled to 2.8T params without
    00:00
    8K