© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34347
    Prev
    Next
    INDEX
    Explanations

    nefarious games, variety of

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

    The neuron is highly sensitive to tokens containing an apostrophe (i.e. possessive ’s or contractions).

    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
    ....)
    -1.04
    صیلات
    -1.02
     diritti
    -0.97
     största
    -0.96
    ...).
    -0.94
     berätt
    -0.93
    Ili
    -0.91
     процессе
    -0.90
     mostrado
    -0.90
     Boletín
    -0.86
    POSITIVE LOGITS
     более
    0.95
    Composable
    0.86
     and
    0.82
    者
    0.81
    beitet
    0.81
    回家
    0.81
    ślę
    0.79
    miş
    0.79
    izeit
    0.78
    hs
    0.78
    Activations Density 0.080%

    No Known Activations