© 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. 577
    Prev
    Next
    INDEX
    Explanations

    This neuron detects descriptive modifiers—typically adjectives or classification words—that introduce a specific variety or subtype in compound names (e.g. “butter” in “butter beans,” “bottle” in “bottle gourd,” “Russian” in “Russian olive,” 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
    ቖ
    0.36
    行程
    0.35
    fordern
    0.34
     акт
    0.34
     জনগণের
    0.33
     субъек
    0.33
    ग्वि
    0.33
     образова
    0.32
    明確
    0.32
    Про
    0.32
    POSITIVE LOGITS
    /
    0.53
     மற்றும்
    0.50
     and
    0.48
     mga
    0.48
     &
    0.48
     as
    0.45
     etc
    0.44
     illetve
    0.44
     agus
    0.43
     maupun
    0.43
    Activations Density 0.000%

    No Known Activations