© 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. 9061
    Prev
    Next
    INDEX
    Explanations

    numbers and associated words like "more" or "over."

    oai_token-act-pair · gemini-2.0-flash

    numerical values or comparisons

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

    Numbers two or three

    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
     dozen
    -1.48
     dozens
    -1.29
    dozen
    -1.13
     dizaines
    -1.00
     hundreds
    -0.93
    zens
    -0.89
     деся
    -0.88
     Hundreds
    -0.88
     decenas
    -0.86
    几十
    -0.85
    POSITIVE LOGITS
     two
    0.73
     three
    0.69
    Three
    0.68
    TWO
    0.68
    two
    0.68
     >=",
    0.65
     Three
    0.63
    Two
    0.63
    trois
    0.60
     TWO
    0.60
    Activations Density 2.983%

    No Known Activations