© 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. 65332
    Prev
    Next
    INDEX
    Explanations

    file operations, data handling, UI, errors

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

    The neuron specializes in spotting numeric literals (tokens composed of digits) in code.

    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
     with
    -2.14
     to
    -2.14
     all
    -1.50
     this
    -1.44
     in
    -1.43
     resulting
    -1.42
     a
    -1.38
     until
    -1.35
     on
    -1.30
    borgen
    -1.24
    POSITIVE LOGITS
     granos
    1.50
     ugyan
    1.38
     mellett
    1.38
    為に
    1.31
     evidentemente
    1.30
     benevol
    1.27
    Similarly
    1.24
     milliers
    1.24
     famosas
    1.23
    様な
    1.23
    Activations Density 0.017%

    No Known Activations