© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Llama3.3-70B-IT
    5. 50-RESID-POST-GF
    7. 4973
    Prev
    Next
    INDEX
    Explanations

    secretly, in private, sneaking out

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.3-70B-IT @ 50-resid-post-gf
    Configuration
    Goodfire/Llama-3.3-70B-Instruct-SAE-l50/Llama-3.3-70B-Instruct-SAE-l50.pt
    Prompts (Dashboard)
    10,000 prompts, 128 tokens each
    Dataset (Dashboard)
    lmsys/lmsys-chat-1m
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .Companion
    -0.10
     unwitting
    -0.10
     teasing
    -0.09
     unint
    -0.09
    rix
    -0.09
    bar
    -0.09
     sport
    -0.08
     заÑģÑĤ
    -0.08
     sit
    -0.08
    ounding
    -0.08
    POSITIVE LOGITS
     secret
    0.39
    secret
    0.31
    ç§ĺ
    0.31
    åģ·
    0.27
     Secret
    0.27
     SECRET
    0.26
    SECRET
    0.26
     behind
    0.25
     ÑĤай
    0.25
     secre
    0.24
    Activations Density 0.155%

    No Known Activations