© 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. 26014
    Prev
    Next
    INDEX
    Explanations

    medical and technical jargon

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

    technical terminology, especially in academic or programming contexts.

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

    The neuron fires on infrequent, domain‐specific tokens—math variable names and symbols, code keywords, or specialized scientific terms.

    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
     and
    -1.41
     when
    -1.19
     these
    -1.05
     while
    -1.03
     When
    -0.97
    /***
    -0.97
     or
    -0.95
    putBoolean
    -0.95
     this
    -0.93
     if
    -0.93
    POSITIVE LOGITS
    烹饪
    1.08
     Schloss
    1.02
    賸
    1.01
    脑子
    0.95
     ridiculously
    0.92
     leſs
    0.90
    栳
    0.90
    螽
    0.89
     harrowing
    0.88
     kinda
    0.86
    Activations Density 0.102%

    No Known Activations