© 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. 24-GEMMASCOPE-2-RES-16K
    7. 1678
    Prev
    Next
    INDEX
    Explanations

    lists of items or concepts

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

    The neuron fires on punctuation tokens (e.g. commas and apostive-s), i.e. standalone punctuation marks.

    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_24_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
    uu
    0.89
    א
    0.86
    ać
    0.84
    vorm
    0.82
    ammo
    0.80
    วม
    0.78
     Tycho
    0.77
    ంబేద్కర్
    0.77
    aña
    0.76
    Topo
    0.76
    POSITIVE LOGITS
     cour
    0.74
    নিবার
    0.71
    </td>
    0.68
     j
    0.67
    mand
    0.67
     compulsory
    0.67
    艰
    0.67
     yourself
    0.66
     prisoners
    0.66
     from
    0.66
    Activations Density 0.000%

    No Known Activations