© 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. 0-GEMMASCOPE-TRANSCODER-16K
    7. 14658
    Prev
    Next
    INDEX
    Explanations

    the word "premium", and also activates (to a lesser extent) keywords associated with being in front in sports or finances.

    oai_token-act-pair · gemini-2.0-flash

    premium

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.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
    ientôt
    -0.85
    enapa
    -0.77
    idenav
    -0.70
    hastly
    -0.67
    原始内容存档于
    -0.67
    oubted
    -0.66
    
    -0.66
    âmara
    -0.65
    escens
    -0.65
    ruptedException
    -0.64
    POSITIVE LOGITS
    <bos>
    1.20
     the
    0.58
     kasarigan
    0.51
     puk
    0.50
    [toxicity=0]
    0.48
    ьаж
    0.47
     potr
    0.47
    term
    0.46
    uxxxx
    0.45
     referenties
    0.45
    Activations Density 1.206%

    No Known Activations