© 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. 5967
    Prev
    Next
    INDEX
    Explanations

    for supporting charges

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

    The neuron is detecting numeric tokens, especially multi‐digit or decimal numbers.

    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
     বিনষ্ট
    0.61
     victim
    0.59
    мани
    0.55
    victim
    0.55
     victims
    0.54
    脆
    0.52
     жерт
    0.52
    ('=
    0.52
    惩
    0.52
     setan
    0.51
    POSITIVE LOGITS
     charges
    1.50
     Charges
    1.41
    charges
    1.37
    Charges
    1.31
     Charge
    1.24
     charge
    1.23
    Charge
    1.16
    charge
    1.15
     accusations
    1.13
     accusation
    1.13
    Activations Density 0.167%

    No Known Activations