© 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. 14045
    Prev
    Next
    INDEX
    Explanations

    pretty common words that do not have much meaning on their own

    oai_token-act-pair · gemini-2.0-flash

    words with specific letter patterns

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

    Old spelling/foreign words

    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
    Демографія
    -0.59
    SourceChecksum
    -0.50
    ніципа
    -0.46
    M
    -0.45
    opress
    -0.44
    <eos>
    -0.43
     "..\..\..\
    -0.43
    mtd
    -0.43
    isesti
    -0.43
     Besøkt
    -0.42
    POSITIVE LOGITS
     Houſe
    0.71
     houſe
    0.71
     Majefty
    0.69
     صوتيه
    0.69
     المعيارى
    0.68
    replyreport
    0.67
     purpoſe
    0.67
    ſelves
    0.66
    wiſe
    0.65
    ſelf
    0.64
    Activations Density 14.429%

    No Known Activations