© 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. 13942
    Prev
    Next
    INDEX
    Explanations

    e.g. c

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

    The neuron fires on content‐heavy verbs and related words expressing actions of thought, ability, or inquiry (e.g. think, can, ensure, asking, already tried).

    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
     współpracy
    0.83
    Референ
    0.82
    ா்
    0.79
    ס
    0.78
    нету
    0.77
    го
    0.76
     grievance
    0.75
     customizations
    0.73
    ج
    0.73
     ću
    0.73
    POSITIVE LOGITS
    er
    1.00
    inh
    0.86
    ter
    0.84
    inen
    0.82
    jar
    0.81
     institu
    0.80
    ear
    0.80
     d
    0.79
    v
    0.79
    eer
    0.79
    Activations Density 0.000%

    No Known Activations