© 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. 3303
    Prev
    Next
    INDEX
    Explanations

    words or phrases that act like discourse markers, such as "more" when used to transition between thoughts.

    oai_token-act-pair · gemini-2.0-flash

    technical specifications

    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
     strongly
    -1.05
     essentially
    -1.05
     definitively
    -1.04
     clearly
    -1.04
     explicitly
    -1.03
     unequivocally
    -1.03
     strictly
    -1.02
     decidedly
    -0.98
     fundamentally
    -0.98
     emphatically
    -0.98
    POSITIVE LOGITS
    KURZBESCHREIBUNG
    0.71
    uxxxx
    0.61
    complexContent
    0.59
     >=",
    0.57
    ValueStyle
    0.56
    })}\
    0.55
    UnusedPrivate
    0.54
    ())));
    0.54
    principalTable
    0.54
     auffi
    0.54
    Activations Density 2.275%

    No Known Activations