© 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. 32998
    Prev
    Next
    INDEX
    Explanations

    equations with numbers and variables

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

    The neuron is selectively detecting HTML numeric character‐reference sequences (i.e. tokens of the form “&#<digits>;”).

    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
     Perſ
    -1.15
     dirait
    -1.06
     thoſe
    -1.06
    らなかった
    -1.02
     his
    -1.01
    后果
    -1.00
    丫头
    -1.00
     tega
    -0.99
    従業
    -0.99
    vaného
    -0.98
    POSITIVE LOGITS
     ساتھ
    1.20
     rendono
    1.15
    protobuf
    1.10
    edakan
    1.10
     with
    1.05
     yetu
    1.05
    DatabaseError
    1.01
    人参
    0.99
     hvilke
    0.97
    électricité
    0.96
    Activations Density 0.003%

    No Known Activations