© 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. 34-GEMMASCOPE-RES-131K
    7. 57357
    Prev
    Next
    INDEX
    Explanations

    database field and record

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

    This neuron fires strongly on numeric tokens—especially standalone numbers and numeric literals in code or SQL snippets.

    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
    ätigung
    -0.89
    omaterials
    -0.78
    uminense
    -0.77
     doğ
    -0.75
     наз
    -0.72
     satchel
    -0.72
    Кре
    -0.68
     Swansea
    -0.68
     sassy
    -0.68
    脚本
    -0.68
    POSITIVE LOGITS
     field
    1.15
     Field
    1.10
    Field
    1.09
     fields
    1.06
    Fields
    1.03
     Fields
    0.98
    フィールド
    0.98
     FIELDS
    0.97
    dbf
    0.96
     record
    0.95
    Activations Density 0.077%

    No Known Activations