© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3539
    Prev
    Next
    INDEX
    Explanations

    between X and Y

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

    This neuron flags unusually rare or surprising tokens (i.e. low‐frequency words or punctuation) that deviate strongly from the model’s typical token distribution.

    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
     explosive
    0.65
     entitles
    0.63
     anda
    0.63
     conveyor
    0.62
     bens
    0.62
     maternity
    0.61
    emergency
    0.60
     onerous
    0.60
    льный
    0.60
    лені
    0.59
    POSITIVE LOGITS
     Θ
    0.69
     Microbiol
    0.65
     (^{
    0.62
     regiones
    0.61
     различи
    0.61
     生産
    0.60
    ımızın
    0.60
    ائص
    0.60
     ビジネス
    0.59
    微生物
    0.59
    Activations Density 0.001%

    No Known Activations