© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 6933
    Prev
    Next
    INDEX
    Explanations

    sequence completion

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

    This neuron consistently lights up on abstract, formal–sounding content words (especially long nominalizations and participles) characteristic of bureaucratic or academic prose.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    을
    0.80
    은
    0.72
    를
    0.71
    ը
    0.71
    BOUNDARY
    0.70
    Ста
    0.70
    StatusBar
    0.66
    ִ
    0.66
    の
    0.65
     το
    0.64
    POSITIVE LOGITS
     ٣
    0.82
     Elektrokhimiya
    0.77
     ١
    0.73
    जिये
    0.69
     Kindern
    0.69
     Einen
    0.69
     zweiten
    0.67
     Vielen
    0.66
     ٢
    0.65
     dunno
    0.65
    Activations Density 0.431%

    No Known Activations