© 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. 15222
    Prev
    Next
    INDEX
    Explanations

    will often find

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

    The neuron primarily activates on the leading words (often capitalized) that begin sentences or major clauses.

    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
     pittores
    -1.08
     gita
    -1.05
     koel
    -1.05
     vru
    -1.05
     jobber
    -1.03
     logis
    -1.03
     ……………………
    -1.02
     hond
    -1.00
     .............
    -1.00
    呛
    -1.00
    POSITIVE LOGITS
     your
    1.98
     you
    1.80
     will
    1.71
     often
    1.41
     many
    1.34
     if
    1.30
     time
    1.28
     sometimes
    1.26
     budete
    1.25
     especially
    1.24
    Activations Density 0.015%

    No Known Activations