© 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. 10-GEMMASCOPE-RES-131K
    7. 1881
    Prev
    Next
    INDEX
    Explanations

    conservatively, conservatorio, conservatives, tonic

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

    The neuron fires on subword tokens starting with “conserv,” i.e. matches the “conserv” root found in words like “conservative,” “conservatively,” or “Conservatism.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    “
    -3.22
    नलोड
    -3.16
    ஂ
    -3.13
     (“
    -2.83
    
    -2.64
    骓
    -2.63
    -“
    -2.56
    鳢
    -2.53
    :”
    -2.53
    沇
    -2.52
    POSITIVE LOGITS
    鉭
    2.63
    enablog
    2.53
     we
    2.47
    躋
    2.42
    翃
    2.39
    estões
    2.34
    買う
    2.33
    僦
    2.31
    騷
    2.27
    薷
    2.25
    Activations Density 0.006%

    No Known Activations