© 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. 12245
    Prev
    Next
    INDEX
    Explanations

    Captain America, double jeopardy, plugins

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

    The neuron responds strongly to capitalized proper names and title‐style words (e.g. movie titles, character names, technical identifiers).

    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
    ций
    -0.88
     Rp
    -0.82
     Pinocchio
    -0.81
     BSD
    -0.79
    UserList
    -0.78
     Waray
    -0.76
    Estonia
    -0.75
     Abar
    -0.74
     Burundi
    -0.73
     Estonia
    -0.73
    POSITIVE LOGITS
     Sk
    1.26
     Carol
    1.02
    Sk
    0.97
     Captain
    0.90
     Brie
    0.89
    Captain
    0.88
    Brie
    0.88
     Binary
    0.88
     Photon
    0.88
     Nick
    0.85
    Activations Density 0.024%

    No Known Activations