© 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. 3745
    Prev
    Next
    INDEX
    Explanations

    `WITH RECURSIVE`

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

    The neuron strongly activates on numeric tokens (and words immediately tagging or introducing them, like “reel”/“reels” before slot-machine reel numbers), so overall it’s detecting numbers and numeric references.

    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
    d
    0.80
     alteration
    0.78
    ly
    0.77
    ae
    0.77
    asus
    0.75
    oelect
    0.75
     woods
    0.74
    died
    0.73
    є
    0.73
     Woods
    0.72
    POSITIVE LOGITS
     Ирина
    0.93
    URLConnection
    0.89
     Современ
    0.86
     いう
    0.85
     сотрудников
    0.84
     einzige
    0.82
     Deshalb
    0.80
     圖
    0.79
     Zusammenarbeit
    0.78
     Екатерина
    0.77
    Activations Density 0.001%

    No Known Activations