© 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. 1790
    Prev
    Next
    INDEX
    Explanations

    component names

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

    The neuron strongly activates on standalone capitalized labels or names that serve as section titles, component/field identifiers, or entity names in technical documentation.

    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
    !
    0.56
    Col
    0.55
    Sol
    0.53
    Mun
    0.51
           
    0.50
    Ча
    0.50
       
    0.50
    sol
    0.50
    Meng
    0.49
    ?
    0.49
    POSITIVE LOGITS
    名の
    0.79
    ごとに
    0.74
    名稱
    0.72
    들의
    0.71
    inizin
    0.70
     names
    0.67
     பெயர்
    0.67
    ઓમાં
    0.65
    们的
    0.64
     이름을
    0.63
    Activations Density 0.000%

    No Known Activations