© 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. 10-GEMMASCOPE-RES-131K
    7. 73809
    Prev
    Next
    INDEX
    Explanations

    contrastive explanations before condition

    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_10/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
     are
    -2.67
    y
    -2.20
    -
    -2.13
    h
    -1.91
    k
    -1.89
     of
    -1.83
    o
    -1.78
    j
    -1.75
    es
    -1.70
    ra
    -1.70
    POSITIVE LOGITS
    ліза
    1.91
     typy
    1.89
     komfort
    1.89
     kupa
    1.88
    ᯾
    1.83
     karton
    1.78
    ……"
    1.74
     unik
    1.74
     koncentr
    1.73
     geografic
    1.72
    Activations Density 0.111%

    No Known Activations