© 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. 6717
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to legal and prison contexts, specifically grievances and violations within a jail setting

    oai_token-act-pair · gemini-2.0-flash

    places, buildings

    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
    GraphicsUnit
    -0.66
    Personendaten
    -0.65
    VersionUID
    -0.57
     panoramique
    -0.54
    oa̍t
    -0.53
    脚注の使い方
    -0.52
     nationaux
    -0.52
    хьтан
    -0.52
     NDEBUG
    -0.51
    Fuckin
    -0.49
    POSITIVE LOGITS
     studio
    1.25
     clinic
    1.15
     shop
    1.13
     restaurant
    1.13
     laboratory
    1.12
     store
    1.11
     stadium
    1.07
     library
    1.04
     hospital
    1.03
    studio
    1.03
    Activations Density 3.824%

    No Known Activations