© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47334
    Prev
    Next
    INDEX
    Explanations

    will be useful

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

    The neuron is selectively firing on software‐license and legal‐boilerplate segments (e.g. copyright notices, version strings, redistribution clauses, “distributed in the hope that it will be useful,” etc.).

    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
    フー
    -0.68
     получил
    -0.68
    ꯡ
    -0.67
     裝
    -0.67
     Lins
    -0.67
    bulos
    -0.65
    ódź
    -0.65
    Temporary
    -0.65
    FFICIENT
    -0.65
     Descriptive
    -0.64
    POSITIVE LOGITS
     שט
    0.77
     című
    0.70
    mies
    0.69
     USER
    0.69
    纓
    0.66
    ᵒ
    0.66
    Capítulo
    0.66
     veterinarian
    0.65
    Penguin
    0.64
    朗
    0.64
    Activations Density 0.049%

    No Known Activations