© 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. 31-GEMMASCOPE-2-RES-16K
    7. 6607
    Prev
    Next
    INDEX
    Explanations

    cute porn, image generation, disabled accounts

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

    The neuron fires strongly on numeric tokens (especially decimals and years) and on the word “sell,” so it’s looking for explicit numbers or sales/transaction terms.

    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_31_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
    ب
    0.83
    ر
    0.81
    𝐬
    0.80
    sa
    0.79
    в
    0.78
    Usuario
    0.77
    sound
    0.76
    ві
    0.75
     uridine
    0.75
    ४
    0.75
    POSITIVE LOGITS
     appoints
    0.75
     kroz
    0.70
     azok
    0.70
     Depends
    0.67
     oike
    0.67
     nemo
    0.66
     afferma
    0.66
     intersects
    0.65
     through
    0.63
     agreements
    0.63
    Activations Density 0.000%

    No Known Activations