© 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. 10-GEMMASCOPE-RES-131K
    7. 71011
    Prev
    Next
    INDEX
    Explanations

    abbreviations and questions

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

    The neuron fires on non-content markers and metadata tokens—things like ellipses, uppercase abbreviations (e.g. state or organizational codes), and license or reference 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_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
    镲
    -2.13
     Ă
    -2.11
     juſ
    -2.09
     eenen
    -2.06
     hectáreas
    -2.05
    𓆏
    -2.03
     raconter
    -1.99
    立马
    -1.98
    阏
    -1.97
    狴
    -1.96
    POSITIVE LOGITS
     in
    2.92
     Despite
    2.70
     This
    2.61
     Although
    2.55
     While
    2.53
     With
    2.53
    This
    2.42
    Despite
    2.42
    When
    2.42
     Even
    2.39
    Activations Density 0.000%

    No Known Activations