© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14922
    Prev
    Next
    INDEX
    Explanations

    unusual states or actions

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

    The neuron primarily fires on rare or distinctive content words—uncommon nouns and verbs that stand out lexically.

    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
     içerisinde
    1.03
     देखील
    1.02
     impactful
    0.95
     devido
    0.94
     vilket
    0.92
     irá
    0.92
     minimalistic
    0.89
     स्वरुप
    0.89
     şekilde
    0.88
    விற்கு
    0.87
    POSITIVE LOGITS
    izdat
    0.84
    iform
    0.79
     वगैरह
    0.75
    iop
    0.69
     драматур
    0.69
    咱們
    0.68
    epoch
    0.67
     heretofore
    0.65
     epoch
    0.64
     बेसिकली
    0.64
    Activations Density 0.007%

    No Known Activations