© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 13910
    Prev
    Next
    INDEX
    Explanations

    tributes, Adams Golf, Grand Prix

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

    The neuron fires strongly on common English function words—pronouns (e.g. he, who), auxiliaries (e.g. have), and their contracted forms (e.g. doesn’t)—often marked by internal punctuation.

    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
     μὲν
    -1.05
     pomoci
    -1.05
     éx
    -1.05
    ומית
    -1.02
     italiens
    -1.02
     Sản
    -1.02
    寒い
    -1.02
     hemd
    -1.02
    itäten
    -1.01
     fokus
    -1.01
    POSITIVE LOGITS
     our
    1.16
     this
    1.09
    ;
    1.05
     my
    0.99
     your
    0.98
     those
    0.98
     add
    0.97
     people
    0.96
     an
    0.93
     A
    0.92
    Activations Density 0.055%

    No Known Activations