© 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. 31-GEMMASCOPE-2-RES-16K
    7. 8903
    Prev
    Next
    INDEX
    Explanations

    tasks and methods

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

    The neuron fires strongly on tokens referring to sensitive data items or credentials—words like “recipe” when naming modules, “keys,” and especially “information.”

    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
     வழங்கப்படும்
    0.67
     предоставля
    0.63
     พร้อม
    0.61
     disponible
    0.58
     включая
    0.58
     плюс
    0.57
     проведен
    0.57
     providing
    0.56
     شدن
    0.56
     تمامی
    0.56
    POSITIVE LOGITS
     horizontally
    0.76
     collaboratively
    0.75
     blindly
    0.75
     unsuccessfully
    0.75
     sideways
    0.70
     autonomously
    0.70
     afuera
    0.68
     manually
    0.68
     illegally
    0.66
     defensively
    0.65
    Activations Density 0.394%

    No Known Activations