© 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. 1975
    Prev
    Next
    INDEX
    Explanations

    offering, depending, with

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

    This neuron fires on boilerplate or formal informational passages—especially ones giving corporate/legal or product‐specification details (e.g. company names, registration numbers, instructions, and descriptive “feature” text).

    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
     fanciful
    0.60
     anzi
    0.58
     riguarda
    0.57
    ೮
    0.56
     amea
    0.56
     boycot
    0.55
     galore
    0.54
     restructure
    0.54
    ೨
    0.54
     exerce
    0.53
    POSITIVE LOGITS
     Необходимо
    0.68
     Prior
    0.66
     Kw
    0.64
     Durant
    0.64
     Naj
    0.60
     Maxim
    0.59
     Ya
    0.58
     Syn
    0.58
     Fil
    0.57
     Vad
    0.57
    Activations Density 0.219%

    No Known Activations