© 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. 9481
    Prev
    Next
    INDEX
    Explanations

    breakdown of steps

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

    The neuron isn’t picking out any particular word class or meaning – it’s simply a positional detector, firing strongly on tokens that occur at the start of new segments/paragraphs (i.e. sentence‐initial tokens).

    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
     indígen
    0.91
    acciones
    0.91
     Sólo
    0.91
     Además
    0.90
    㕸
    0.89
     Некоторые
    0.85
     También
    0.85
    ীবন
    0.85
    ieva
    0.84
     Ayrıca
    0.83
    POSITIVE LOGITS
    值
    0.70
     heml
    0.67
     बचा
    0.63
     fun
    0.63
     mysterious
    0.61
    分辨
    0.61
    рас
    0.61
    甭
    0.61
    п
    0.60
     ")");
    0.60
    Activations Density 0.000%

    No Known Activations