© 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. 31-GEMMASCOPE-2-RES-16K
    7. 5994
    Prev
    Next
    INDEX
    Explanations

    cell columns

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

    The neuron consistently activates on numeric tokens (e.g. coordinates, parameter values, line numbers, measurements) and thus is looking for numbers embedded in technical or code‐like text.

    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
    vrdr
    0.62
    耇
    0.61
     ingenuity
    0.59
    鲮
    0.59
     impeachment
    0.59
     разными
    0.59
     oamen
    0.58
     oameni
    0.55
    鷸
    0.55
     জনগণকে
    0.54
    POSITIVE LOGITS
    offset
    0.86
    Offset
    0.73
     Offset
    0.72
     midway
    0.69
     offset
    0.69
    偏移
    0.68
    addContainerGap
    0.67
     halfway
    0.66
     offsets
    0.66
    setLocation
    0.66
    Activations Density 0.034%

    No Known Activations