© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 48484
    Prev
    Next
    INDEX
    Explanations

    compound word and phrase beginnings

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     after
    -1.01
     one
    -0.98
     вза
    -0.87
     goes
    -0.83
     took
    -0.81
     seems
    -0.79
     Queen
    -0.78
     serves
    -0.75
     once
    -0.75
     what
    -0.75
    POSITIVE LOGITS
    Tarif
    0.82
    私も
    0.79
    が見えて
    0.79
     künstler
    0.76
     Erg
    0.75
     pilla
    0.75
    お母
    0.75
     négociations
    0.75
     desinf
    0.75
     asian
    0.75
    Activations Density 0.000%

    No Known Activations