© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1593
    Prev
    Next
    INDEX
    Explanations

    conversational phrases and entities

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

    The neuron strongly activates on the punctuation‐and‐digit pattern of a sports scoreline (e.g. “4-0”).

    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
    habitation
    -0.91
    aqu
    -0.82
    acta
    -0.79
    ΑΙ
    -0.79
    omenclature
    -0.75
     Kro
    -0.75
     warszaw
    -0.73
    QueryParam
    -0.71
     indo
    -0.68
    achat
    -0.68
    POSITIVE LOGITS
     gada
    0.85
     alpina
    0.81
     hdi
    0.71
     Schicks
    0.71
     Sabtu
    0.71
     Wings
    0.69
     ♂
    0.69
     pajak
    0.68
     braz
    0.68
     yake
    0.68
    Activations Density 0.020%

    No Known Activations