© 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. 13973
    Prev
    Next
    INDEX
    Explanations

    Australian treaty

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

    This neuron spikes on the few infrequent or out-of-vocabulary subword tokens (e.g. “different,” “sales,” “(,” “imply”) that are much rarer than the surrounding, more common words.

    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
     outcrops
    0.81
     piante
    0.80
     getBitmap
    0.77
     ornaments
    0.77
     hybr
    0.74
     moldings
    0.73
    ോഗ
    0.71
    h
    0.71
    gur
    0.70
     milliliters
    0.70
    POSITIVE LOGITS
    あえず
    0.64
     recue
    0.64
    가능
    0.63
     NÃO
    0.62
    セプト
    0.61
    不久
    0.61
     mandated
    0.60
    不懂
    0.60
    ção
    0.59
     Política
    0.59
    Activations Density 0.001%

    No Known Activations