© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 7006
    Prev
    Next
    INDEX
    Explanations

    for the benefit of

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

    The neuron is detecting formal “for the benefit of …” constructions—i.e. occurrences of the words “benefit,” “of,” and “the” in that legal‐style phrase.

    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
     reinforces
    -1.54
     ironically
    -1.53
     at
    -1.53
     from
    -1.52
     on
    -1.51
     emphasizes
    -1.44
     promotes
    -1.35
     other
    -1.34
     both
    -1.34
     destacado
    -1.34
    POSITIVE LOGITS
    really
    1.66
     ול
    1.55
     inſ
    1.51
     dismal
    1.48
     judicia
    1.48
    b
    1.48
    那些
    1.46
    j
    1.46
    LY
    1.46
    mostly
    1.45
    Activations Density 0.036%

    No Known Activations