© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 13187
    Prev
    Next
    INDEX
    Explanations

    mentions of mistreatment or violence toward people based on group identity, often including the words 'beaten' or 'attacked'.

    oai_token-act-pair · gemini-2.0-flash

    Descriptions of violence

    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
    Autoritní
    -0.59
    TagMode
    -0.56
     autorytatywna
    -0.55
    getDescriptor
    -0.48
    apter
    -0.47
    Rhestr
    -0.47
    Datuak
    -0.47
    justed
    -0.46
    utz
    -0.46
    
    -0.46
    POSITIVE LOGITS
     lenker
    0.62
    InitVars
    0.62
    ]")]
    0.60
    awaiter
    0.58
    يميديا
    0.57
     ویکی‌آمباردا
    0.56
    WithIOException
    0.56
    Kilder
    0.52
    %)$
    0.51
     المعيارى
    0.51
    Activations Density 10.077%

    No Known Activations