© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 2836
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on words that introduce or label list‐style or section headings (e.g. “Guest,” “stars,” “Other,” “support,” “pieces,” “products,” etc.).

    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
    जब
    0.64
     hecha
    0.61
    ときに
    0.60
     "**
    0.59
     unidirectional
    0.59
    だから
    0.58
     undirected
    0.57
    ^(-
    0.57
     verdadera
    0.57
     obligación
    0.57
    POSITIVE LOGITS
     include
    1.69
    include
    1.44
     includes
    1.43
     Include
    1.41
     incluem
    1.31
     Includes
    1.30
     INCLUDE
    1.23
     incluyen
    1.22
     meliputi
    1.21
     включают
    1.20
    Activations Density 0.164%

    No Known Activations