© 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. 11158
    Prev
    Next
    INDEX
    Explanations

    used for any illegal, harmful, violent, racist, or sexual purposes

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

    The neuron is primarily activated by standalone numeric tokens (i.e. numbers or number-like strings).

    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
     hinzufügen
    0.57
    ősz
    0.56
     baptized
    0.56
    ೀಯ
    0.55
     ILL
    0.54
    шки
    0.53
     धर्म
    0.53
    িলার
    0.53
     batman
    0.53
    Ill
    0.52
    POSITIVE LOGITS
     HN
    0.66
    HN
    0.65
     Hack
    0.57
     hn
    0.57
    Employees
    0.57
    hn
    0.55
     Ar
    0.55
    现代化
    0.54
    Hack
    0.53
     FH
    0.53
    Activations Density 0.026%

    No Known Activations