© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 39281
    Prev
    Next
    INDEX
    Explanations

    digital content and media types

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

    The neuron lights up on concrete noun targets of actions (direct objects)—for example “files,” “images,” “content,” “bugs,” etc.—i.e. things being acted upon.

    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
    ținut
    -1.06
     wszyscy
    -0.97
     Hauses
    -0.96
    AGAIN
    -0.95
    ὲν
    -0.93
     모든
    -0.93
     everyone
    -0.93
     omdat
    -0.93
    因为
    -0.90
    parentElement
    -0.90
    POSITIVE LOGITS
     via
    1.95
     through
    1.79
     or
    1.62
     yourself
    1.54
     you
    1.50
     by
    1.50
     without
    1.37
     using
    1.35
     if
    1.34
     your
    1.31
    Activations Density 0.072%

    No Known Activations