© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11392
    Prev
    Next
    INDEX
    Explanations

    python code definitions

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

    This neuron detects the imperative‐style action verbs that open commit or change‐log entries (e.g. “Add,” “Fix,” “Rename,” “Rewrite,” “Support,” 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_31_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
     recessed
    0.60
    이고
    0.52
     Sudan
    0.51
    㧿
    0.51
     Conse
    0.51
    ffer
    0.51
    ুঁ
    0.50
    צות
    0.50
     ibn
    0.50
    conse
    0.50
    POSITIVE LOGITS
     __
    2.23
    .__
    2.05
    __
    1.90
     "__
    1.73
    ).__
    1.65
    ___
    1.64
     '__
    1.63
     ___
    1.59
    ().__
    1.57
     (__
    1.48
    Activations Density 0.005%

    No Known Activations