© 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. 38720
    Prev
    Next
    INDEX
    Explanations

    fully described or available

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

    The neuron fires on occurrences of the string “full” (e.g. “full,” “fully,” “fully-described”), i.e. it detects variants of the word “full.”

    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
     then
    -1.85
    橹
    -1.63
    銲
    -1.59
    杙
    -1.55
    Ң
    -1.53
     annet
    -1.51
    ANTE
    -1.49
     denoting
    -1.48
    탉
    -1.48
    嗵
    -1.45
    POSITIVE LOGITS
    過ぎる
    1.92
    While
    1.88
    On
    1.73
    Not
    1.65
    That
    1.62
    2
    1.60
    If
    1.59
    Though
    1.58
     этого
    1.58
    過來
    1.57
    Activations Density 0.023%

    No Known Activations