© 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. 22-GEMMASCOPE-TRANSCODER-16K
    7. 12808
    Prev
    Next
    INDEX
    Explanations

    descriptive text that relates a story with strong human emotion and/or imagery

    oai_token-act-pair · gemini-2.0-flash

    shining

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
     ſeveral
    -0.81
     purpoſe
    -0.80
     Efq
    -0.76
     Majefty
    -0.72
    setVerticalGroup
    -0.72
     becauſe
    -0.71
    ImageContext
    -0.70
     يتيمه
    -0.69
     reaſon
    -0.69
     beſt
    -0.69
    POSITIVE LOGITS
     referrerpolicy
    0.43
    oxin
    0.42
    tagHelper
    0.41
    czy
    0.41
     upo
    0.41
     hand
    0.40
    erd
    0.39
    peteer
    0.39
     process
    0.39
     reb
    0.39
    Activations Density 0.885%

    No Known Activations