© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66543
    Prev
    Next
    INDEX
    Explanations

    concepts and their attributes

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

    The neuron flags formal multi-word noun phrases—especially technical or structural compound nouns (e.g., “dosage schedule,” “error format,” “display technology”).

    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
     světě
    -1.77
    throughout
    -1.63
    usions
    -1.57
    weren
    -1.48
    wtf
    -1.47
     bílé
    -1.46
    indicates
    -1.44
     
    -1.43
    Юлия
    -1.43
    ſche
    -1.43
    POSITIVE LOGITS
     of
    3.94
     is
    3.84
     was
    3.28
     that
    2.20
     will
    2.17
     are
    2.05
     must
    2.03
     always
    1.94
    は
    1.92
     would
    1.89
    Activations Density 0.168%

    No Known Activations