© 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. 33149
    Prev
    Next
    INDEX
    Explanations

    .cpp files and units

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

    The neuron is primarily triggered by numerical tokens (e.g. standalone numbers).

    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
     Despite
    -1.47
     While
    -1.41
     (
    -1.41
     despite
    -1.37
     from
    -1.31
     an
    -1.31
     Although
    -1.30
     because
    -1.30
     ((
    -1.27
    Despite
    -1.27
    POSITIVE LOGITS
     maniere
    1.33
     and
    1.27
     crede
    1.23
     traged
    1.20
     professione
    1.17
     uſe
    1.15
     mense
    1.14
     ladri
    1.13
     discrepan
    1.10
     monaster
    1.10
    Activations Density 0.001%

    No Known Activations