© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 11453
    Prev
    Next
    INDEX
    Explanations

    consult with

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

    The neuron selectively fires on text describing roofing repair services and related promotional language (e.g. offers to identify, resolve, or prevent roof damage).

    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
    一下子
    0.77
     undersea
    0.69
     nearly
    0.66
     duniya
    0.66
     iki
    0.65
     zindagi
    0.65
     things
    0.65
     ilk
    0.64
     spawned
    0.64
     unicode
    0.63
    POSITIVE LOGITS
    シンプルな
    0.84
    𝗛
    0.84
     бренда
    0.82
    óloga
    0.82
    angebot
    0.81
    𝙖
    0.81
    優れた
    0.80
     magní
    0.79
     しっかり
    0.79
    irà
    0.79
    Activations Density 0.408%

    No Known Activations