© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 15005
    Prev
    Next
    INDEX
    Explanations

    fluid retention, blood flow, oxygen diffusion

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

    technical terms in scientific, technical, or programming documentation.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron strongly activates on domain‐specific technical or scientific terms (e.g. jargon‐style nouns).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     tecnológica
    -0.87
     ceva
    -0.82
     obi
    -0.82
    顺着
    -0.82
    ケモン
    -0.81
     creato
    -0.79
     画面
    -0.77
     longtemps
    -0.76
    を持っている
    -0.76
    uhan
    -0.75
    POSITIVE LOGITS
     and
    0.91
    ;/
    0.84
    naments
    0.82
    ΕΙ
    0.80
    letons
    0.80
     rodillas
    0.79
     lluvias
    0.79
    xbb
    0.79
    藐
    0.78
    ));
    0.77
    Activations Density 0.112%

    No Known Activations