© 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. 38752
    Prev
    Next
    INDEX
    Explanations

    credit, place names, cms, kcal

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

    The neuron fires strongly on isolated domain-specific keywords and abbreviations (e.g. “cred,” “cms,” “kcal,” “Heidelberg”), i.e. technical or specialized noun tokens.

    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
    胳
    -2.48
     nuestros
    -2.41
    ANIEL
    -2.36
    轳
    -2.34
     Gestão
    -2.33
    噚
    -2.30
    缉
    -2.28
     stratég
    -2.28
    rierten
    -2.20
    だが
    -2.14
    POSITIVE LOGITS
    的
    3.58
    s
    3.25
    am
    3.20
    ay
    2.59
    k
    2.55
    a
    2.55
    2
    2.38
    section
    2.36
    of
    2.31
    up
    2.31
    Activations Density 0.004%

    No Known Activations