© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12766
    Prev
    Next
    INDEX
    Explanations

    rock music or Rocky Linux

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

    The neuron strongly activates on occurrences of the substring “rock” (e.g. “Rock,” “rock,” or compounds beginning with “rock”).

    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
    褒
    0.57
    /**
    0.53
    霧
    0.52
    мов
    0.51
     भत्ता
    0.50
    云
    0.50
    ۴
    0.50
    »
    0.50
     cloud
    0.50
    ности
    0.50
    POSITIVE LOGITS
    Rock
    0.99
    efeller
    0.98
     Rock
    0.94
    石头
    0.93
    ROCK
    0.89
     ROCK
    0.89
    Rocks
    0.89
     outcrop
    0.87
     outcrops
    0.86
     rok
    0.84
    Activations Density 0.125%

    No Known Activations