© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 81086
    Prev
    Next
    INDEX
    Explanations

    software class/component names

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

    The neuron fires on the key technical nouns and names in code comment headers—words like “compressor,” “filter,” “implementation,” “module,” “script,” etc., that describe a file or component’s purpose.

    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
     ansiedad
    -1.30
    事も
    -1.30
     véhic
    -1.25
     huellas
    -1.23
     lluvias
    -1.18
     anglick
    -1.17
     sustancias
    -1.16
     citada
    -1.14
     regalías
    -1.13
    ictured
    -1.13
    POSITIVE LOGITS
     for
    2.08
     that
    1.51
     для
    1.45
     used
    1.42
    。(
    1.37
     with
    1.32
    。
    1.32
    .
    
    1.28
    ।
    1.18
    .<
    1.15
    Activations Density 0.145%

    No Known Activations