© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 39729
    Prev
    Next
    INDEX
    Explanations

    capitalized words or start of words

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron flags content‐bearing “key” words—especially relatively uncommon or domain‐specific terms (proper nouns, technical words, etc.)—in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     of
    -2.05
     the
    -1.73
     that
    -1.59
     vervolgens
    -1.57
     avevano
    -1.43
    然后
    -1.41
     it
    -1.39
    gebühr
    -1.36
     either
    -1.34
     hadde
    -1.33
    POSITIVE LOGITS
     new
    1.78
    They
    1.71
    While
    1.63
    </h2>
    1.62
     recientes
    1.52
    Perhaps
    1.52
     olduğu
    1.50
     name
    1.49
     они
    1.48
    Sometimes
    1.47
    Activations Density 0.075%

    No Known Activations