© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129970
    Prev
    Next
    INDEX
    Explanations

    The neuron is sensitive to proper names and branded or specialized technical terms (i.e. capitalized entity tokens).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    compound words and phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    が
    -1.60
    を
    -1.29
    2
    -1.25
    6
    -1.24
    8
    -1.20
    umán
    -1.16
     nobili
    -1.16
    こういった
    -1.13
    <bos>
    -1.10
     adul
    -1.10
    POSITIVE LOGITS
     godk
    1.26
     hó
    1.23
    ônus
    1.23
     öppna
    1.21
     verksamhet
    1.21
     necesit
    1.20
    ᵉ
    1.18
    ヱ
    1.17
     secado
    1.15
     vanligt
    1.13
    Activations Density 0.096%

    No Known Activations