© 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. 72047
    Prev
    Next
    INDEX
    Explanations

    words like partially, moderately, slightly

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

    The neuron fires on semantically rich content words—especially substantive nouns and other key descriptive terms (e.g. people’s roles, proper names, technical 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
    particularly
    -1.27
    thousands
    -1.16
    mainly
    -1.16
    hundreds
    -1.14
     bestellen
    -1.11
    锨
    -1.05
    ticularly
    -1.05
    ſſen
    -1.05
    primarily
    -1.05
    耵
    -1.02
    POSITIVE LOGITS
     partially
    1.19
     two
    1.17
     mildly
    1.13
     moderately
    1.12
     slightly
    1.09
     TWO
    1.09
     three
    1.06
     subtly
    1.00
     one
    1.00
     Although
    0.99
    Activations Density 0.119%

    No Known Activations