© 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. 67641
    Prev
    Next
    INDEX
    Explanations

    sequence of characters or words

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

    The neuron strongly activates on repeated occurrences of the uppercase acronym “PWN,” i.e. sequences of the same three‐letter uppercase token.

    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
    istolet
    -0.97
     from
    -0.97
     JUST
    -0.96
     necessarie
    -0.95
     interessieren
    -0.95
     滅
    -0.94
     diligent
    -0.93
     vestido
    -0.91
     or
    -0.91
    There
    -0.90
    POSITIVE LOGITS
     teat
    1.06
    也被
    0.93
     această
    0.93
     禁止
    0.92
    甚至是
    0.92
     centimeter
    0.92
     qt
    0.90
     dee
    0.90
    nefs
    0.90
    anej
    0.88
    Activations Density 0.053%

    No Known Activations