© 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. 19-GEMMASCOPE-TRANSCODER-16K
    7. 2156
    Prev
    Next
    INDEX
    Explanations

    This neuron seems to activate on seemingly random words and punctuation in technical and medical texts, but it seems to have a bias towards words with numerical values

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_19/width_16k/average_l0_12
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.19.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
    <bos>
    -0.56
     designations
    -0.52
     voices
    -0.52
    EndContext
    -0.52
     brains
    -0.49
    RectangleBorder
    -0.49
     designation
    -0.49
     animals
    -0.49
    CGRectMake
    -0.47
    Дан
    -0.46
    POSITIVE LOGITS
    tvguidetime
    0.60
     permanent
    0.60
    __(/*!
    0.59
    Архівовано
    0.57
    FirstResponder
    0.57
     incomplète
    0.55
    "]["
    0.54
     विश्वसनीय
    0.54
     blessés
    0.53
     durable
    0.52
    Activations Density 9.887%

    No Known Activations