© 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. 10-GEMMASCOPE-RES-131K
    7. 19394
    Prev
    Next
    INDEX
    Explanations

    instructions and questions

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

    The neuron fires on instruction or question prompt keywords (e.g. “List,” “What,” “Calculate,” “Do,” etc.) that begin a request or command.

    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
    /
    -2.28
    :
    -2.20
    _
    -2.02
    -
    -1.98
     —
    -1.66
     an
    -1.55
    —
    -1.54
     or
    -1.50
    --
    -1.44
     I
    -1.42
    POSITIVE LOGITS
    igem
    1.64
     ſte
    1.63
    萏
    1.59
    嫑
    1.57
    鹇
    1.56
    geschenk
    1.55
    formatie
    1.52
    volve
    1.52
     fallu
    1.51
    issent
    1.50
    Activations Density 0.026%

    No Known Activations