© 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. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 9250
    Prev
    Next
    INDEX
    Explanations

    phrases in which the word "what" is followed or preceded by a personal pronoun

    oai_token-act-pair · gemini-2.0-flash

    phrases involving personal significance or desire

    np_token-act-pair-logits · gpt-4o-mini

    what

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NUMX
    -0.93
     ་་
    -0.81
    crapers
    -0.81
     InputDecoration
    -0.80
    InitVars
    -0.79
     ―――――
    -0.74
     vectorstock
    -0.73
     Paglinawan
    -0.73
    TagHelper
    -0.71
    itempty
    -0.71
    POSITIVE LOGITS
     matters
    0.61
     doesn
    0.57
     does
    0.57
     is
    0.52
     comes
    0.51
    sim
    0.47
     it
    0.47
     MATTERS
    0.46
    tal
    0.46
    el
    0.45
    Activations Density 1.109%

    No Known Activations