© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 30822
    Prev
    Next
    INDEX
    Explanations

    where or where clauses

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

    The neuron flags uncommon, content‐bearing words (rare or domain‐specific terms) rather than frequent function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    \}=
    -1.14
    encies
    -1.11
    之上
    -1.10
    getCause
    -1.06
     quale
    -1.06
    ksiä
    -1.05
    riceps
    -1.03
    绻
    -1.02
    ˾
    -1.02
    нологи
    -1.02
    POSITIVE LOGITS
     where
    12.81
    where
    7.88
    Where
    7.00
     где
    6.88
     Where
    6.59
     donde
    6.53
     où
    6.19
     gdzie
    5.91
     WHERE
    5.91
     حيث
    5.28
    Activations Density 1.462%

    No Known Activations