© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 37709
    Prev
    Next
    INDEX
    Explanations

    copyright and license

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

    Explanation of neuron 4 behavior: the main thing this neuron does is find numeric tokens (e.g. version numbers, floats, or other digit sequences).

    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
    ulsory
    -0.68
     isomer
    -0.66
     Albany
    -0.66
    afio
    -0.65
     nullable
    -0.65
    MEP
    -0.65
     Lizzie
    -0.65
    itored
    -0.65
    İY
    -0.64
    prudence
    -0.63
    POSITIVE LOGITS
    conditions
    0.78
    icken
    0.71
     Hal
    0.71
    itaria
    0.66
    ividual
    0.66
     violets
    0.64
     dominate
    0.64
     künstler
    0.64
     Zab
    0.64
    vai
    0.63
    Activations Density 0.064%

    No Known Activations