© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1694
    Prev
    Next
    INDEX
    Explanations

    code keywords like try, print, if

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

    The neuron strongly activates on schema‐style or specification identifiers—especially quoted multiword names or uppercase field/parameter labels.

    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
    Caf
    0.60
    É
    0.60
    Cómo
    0.59
    Alk
    0.59
    Notre
    0.58
    Jean
    0.58
    Étienne
    0.57
     林
    0.57
    Laur
    0.57
    không
    0.57
    POSITIVE LOGITS
     formatted
    0.44
     jumlah
    0.43
     presenceData
    0.41
     desired
    0.41
     initialize
    0.41
     offsetX
    0.41
     covariate
    0.41
     varsa
    0.41
     bize
    0.40
     pretend
    0.40
    Activations Density 0.004%

    No Known Activations