© 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. 6680
    Prev
    Next
    INDEX
    Explanations

    specify the ID

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

    This neuron detects the first content‐bearing token at the start of each new segment (i.e. the very first “meaningful” word after a <start>).

    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
     enteral
    0.92
     arrhyth
    0.91
     Epile
    0.84
     Beige
    0.84
     epilept
    0.84
    j
    0.84
     moisturizing
    0.83
     itibaren
    0.83
     producir
    0.82
     antise
    0.82
    POSITIVE LOGITS
    نة
    0.76
    likle
    0.74
    ضاع
    0.71
    橼
    0.68
    лює
    0.68
    clusions
    0.68
    લું
    0.66
    brella
    0.66
    gian
    0.65
    atakse
    0.64
    Activations Density 0.000%

    No Known Activations