© 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. 34111
    Prev
    Next
    INDEX
    Explanations

    numbers and codes

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

    The neuron activates on numeric tokens—i.e. sequences of digits such as train numbers, case or section numbers, dates, and other multi‐digit identifiers.

    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
    INCLUDING
    -1.27
     nacidos
    -1.25
    ところに
    -1.18
    ₿
    -1.18
    𝟏
    -1.17
    
    -1.17
    になるので
    -1.16
     yürü
    -1.13
     웹
    -1.13
     uſ
    -1.12
    POSITIVE LOGITS
    1
    1.93
    2
    1.74
    5
    1.67
    3
    1.66
     these
    1.64
    7
    1.56
    4
    1.53
    9
    1.37
    <bos>
    1.36
    6
    1.35
    Activations Density 0.041%

    No Known Activations