© 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. 67534
    Prev
    Next
    INDEX
    Explanations

    things, parameter, ribs, value, map, variable

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

    The neuron fires on technical jargon and domain‐specific nouns and acronyms (e.g. “Internet of Things,” “parameter,” “map,” “ribs”) indicative of a highly technical context.

    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
     mosso
    -1.30
     absur
    -1.30
     attent
    -1.27
     gymnas
    -1.22
    ने
    -1.18
     horri
    -1.17
    taient
    -1.17
     departe
    -1.17
     gekocht
    -1.17
     produs
    -1.16
    POSITIVE LOGITS
     but
    1.91
     a
    1.83
     more
    1.77
     another
    1.70
     if
    1.51
     better
    1.38
     other
    1.36
     many
    1.36
     can
    1.35
     both
    1.35
    Activations Density 0.193%

    No Known Activations