© 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. 34-GEMMASCOPE-RES-131K
    7. 23537
    Prev
    Next
    INDEX
    Explanations

    input related labels

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

    The neuron fires on literal values or identifiers that are being assigned or passed as arguments—i.e. the tokens immediately following “=” signs or argument‐keywords in code or markup.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    signor
    -0.88
    nees
    -0.83
    itemId
    -0.79
    😼
    -0.77
     للاست
    -0.77
    asses
    -0.75
    ільки
    -0.74
     Arhi
    -0.74
    】：
    -0.74
    unie
    -0.73
    POSITIVE LOGITS
    helper
    1.12
     Input
    1.08
     full
    1.08
     variant
    1.02
     helper
    0.96
    HEL
    0.95
    Input
    0.94
     margin
    0.92
    Variant
    0.92
     input
    0.90
    Activations Density 0.000%

    No Known Activations