© 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. 3122
    Prev
    Next
    INDEX
    Explanations

    Ten/Tin/Tensor/Tant

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

    The neuron fires on technical size or model codes—particularly tokens like “T25,” “T20,” etc., that label specific tool or bit sizes.

    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
    ثة
    0.69
    izion
    0.57
    psies
    0.56
    者を
    0.55
    豉
    0.55
     מו
    0.54
     neve
    0.54
     Dubey
    0.54
     mecanismo
    0.54
    ิทธิภาพ
    0.54
    POSITIVE LOGITS
     T
    0.76
     टी
    0.71
     Tind
    0.71
    Shirts
    0.68
     Tot
    0.67
    恤
    0.66
     Т
    0.63
     Titans
    0.63
    sunami
    0.63
     Tant
    0.63
    Activations Density 0.268%

    No Known Activations