© 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. 15144
    Prev
    Next
    INDEX
    Explanations

    the word "if" along with other discourse-related words that express opinions as well as location names. It also finds the non-auxiliary verb "have" and some first-person-view words

    oai_token-act-pair · gemini-2.0-flash

    Anyone

    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
     חיצוניים
    -0.57
    corbic
    -0.56
    isat
    -0.55
     Dummies
    -0.54
    BarStyle
    -0.53
    createFrom
    -0.53
     fread
    -0.52
    ///</
    -0.52
    eens
    -0.50
    something
    -0.49
    POSITIVE LOGITS
     anyone
    2.61
     anything
    2.25
    anyone
    2.16
     anybody
    2.16
     Anyone
    2.11
    Anyone
    2.05
    anything
    1.97
     Anything
    1.94
    Anything
    1.92
     Anybody
    1.91
    Activations Density 5.289%

    No Known Activations