© 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. 22-GEMMASCOPE-RES-131K
    7. 64974
    Prev
    Next
    INDEX
    Explanations

    line or row

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

    The neuron is primarily triggered by programming‐and‐UI-related jargon—especially tokens describing user‐interface events and controls (e.g. “button,” “popup,” “window,” “click,” “load,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ことも
    -1.44
    He
    -1.33
    為に
    -1.27
    licopter
    -1.26
    R
    -1.25
    を作って
    -1.24
     zweifel
    -1.19
    韌
    -1.19
    E
    -1.18
    </em>
    -1.17
    POSITIVE LOGITS
     there
    2.95
     we
    2.73
     it
    2.63
     you
    2.61
     they
    2.61
    there
    1.51
    釧
    1.49
     هناك
    1.48
    を見
    1.44
    があります
    1.44
    Activations Density 0.199%

    No Known Activations