© 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. 13020
    Prev
    Next
    INDEX
    Explanations

    balance, short, value, choice

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

    The neuron strongly activates on tokens that start a new paragraph or section (i.e. the first word immediately following a blank‐line break).

    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
     certe
    0.86
    此事
    0.86
    ן
    0.80
     ador
    0.79
     Populate
    0.79
     attacc
    0.77
    つけて
    0.76
     există
    0.76
     ござい
    0.75
    jela
    0.74
    POSITIVE LOGITS
    ymmetry
    0.77
     ferrous
    0.72
    Scot
    0.68
    arsipkan
    0.68
    齑
    0.66
    RIC
    0.66
     السي
    0.66
    dominated
    0.66
     SCOTT
    0.65
     paces
    0.64
    Activations Density 0.001%

    No Known Activations