© 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. 13158
    Prev
    Next
    INDEX
    Explanations

    palindrome

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

    The neuron activates on hidden or zero‐width formatting tokens (invisible layout/position markers) rather than normal 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
     arithmetic
    0.80
     attenuation
    0.77
     Ư
    0.73
     underestimate
    0.73
     Pythagorean
    0.72
     tensor
    0.71
     contaminant
    0.71
     cargar
    0.71
     subplot
    0.71
     token
    0.70
    POSITIVE LOGITS
    いた
    0.80
    <0xD7>
    0.76
     وبه
    0.72
    YL
    0.71
    सेज
    0.69
    lerden
    0.69
    并
    0.68
    вшей
    0.67
    Baxter
    0.67
    릭
    0.66
    Activations Density 0.008%

    No Known Activations