© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 14026
    Prev
    Next
    INDEX
    Explanations

    strong commitment or value

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

    The neuron activates on words signaling emphasis or significance—particularly adjectives and phrases like “important,” “forgotten,” or “if any” that mark something as noteworthy.

    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
     paramètres
    0.88
    atiquement
    0.79
    regulating
    0.79
    cier
    0.79
     capelli
    0.78
    ক্কা
    0.77
    eningkatan
    0.77
    ={<
    0.76
     centerline
    0.76
    থাৎ
    0.75
    POSITIVE LOGITS
    れば
    0.88
    ಮಾ
    0.80
     LISTEN
    0.79
    TA
    0.78
    レント
    0.78
    ᅴ
    0.77
     KMnO
    0.74
     Ory
    0.74
    ری
    0.73
    ια
    0.73
    Activations Density 0.000%

    No Known Activations