© 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. 75185
    Prev
    Next
    INDEX
    Explanations

    letting things be

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

    The neuron responds to content‐bearing words—mostly mid‐frequency nouns and verbs that carry substantial meaning rather than to grammatical function words.

    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
    λιο
    -0.88
     Manual
    -0.88
     músico
    -0.86
    memoized
    -0.81
     favorito
    -0.80
     müs
    -0.79
     lég
    -0.79
     manuales
    -0.79
    差し
    -0.78
     alimentare
    -0.77
    POSITIVE LOGITS
     your
    1.05
     dictate
    1.04
    stdio
    0.93
    )):
    0.88
     dictates
    0.86
    océan
    0.85
    PositiveButton
    0.85
    優先
    0.85
     their
    0.84
    ניים
    0.84
    Activations Density 0.029%

    No Known Activations