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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 8-GEMMASCOPE-TRANSCODER-16K
    7. 71
    Prev
    Next
    INDEX
    Explanations

    It's really hard to say, but the neuron seems to activate on a wide variety of proper nouns and commands, possibly preferring capitalized words or command line instructions or variable names.

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_8/width_16k/average_l0_52
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.8.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
    Tikang
    -0.78
     виправивши
    -0.76
     يتيمه
    -0.71
     насељу
    -0.69
     Administrativna
    -0.65
    TintMode
    -0.60
    cracker
    -0.59
    postmedia
    -0.57
    pressor
    -0.54
    Panamoan
    -0.54
    POSITIVE LOGITS
    ViewFeatures
    0.56
     uman
    0.54
    >{@
    0.48
     quoque
    0.48
     habet
    0.47
    uman
    0.47
     mentale
    0.44
     setSelected
    0.44
    enumerable
    0.43
     presentato
    0.43
    Activations Density 8.427%

    No Known Activations