© 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. 120
    Prev
    Next
    INDEX
    Explanations

    **describing complex features**

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

    positive adjectives used to describe quality or value.

    oai_token-act-pair · claude-4-5-haikuTriggered by @jyhe0408

    The neuron fires on strongly positive, persuasive descriptive adjectives and buzzwords used in marketing or promotional language.

    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
    0.55
     that
    0.54
    ()
    0.54
    .
    0.51
     the
    0.46
    ,
    0.45
    й
    0.44
     from
    0.43
    0
    0.43
    ↵↵
    0.43
    POSITIVE LOGITS
     personable
    0.72
     unapolog
    0.70
     uncompromising
    0.68
     adaptable
    0.68
     multifaceted
    0.67
     নির্ভরযোগ্য
    0.59
     scalable
    0.58
    <unused615>
    0.57
     soulful
    0.56
     enviable
    0.56
    Activations Density 0.650%

    No Known Activations