© 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. 889
    Prev
    Next
    INDEX
    Explanations

    GPT followed by hyphen and number

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

    This neuron activates on technical identifiers combining uppercase letters with digits—e.g. media codecs and container formats (MPEG2, MP4, DivX), protocol/version labels (IP version 4/6), or similar alphanumeric codes.

    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
    <unused3>
    0.80
    <unused21>
    0.75
    <unused29>
    0.73
    <unused2>
    0.72
    <unused546>
    0.72
     dokładnie
    0.71
    <unused89>
    0.70
    <unused43>
    0.68
    <unused92>
    0.66
    <unused42>
    0.65
    POSITIVE LOGITS
    III
    1.27
    II
    1.20
     III
    1.15
     II
    1.12
    2
    1.05
    ®
    1.05
    1
    1.05
    3
    1.00
    4
    0.94
    (
    0.92
    Activations Density 0.555%

    No Known Activations