© 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. 72524
    Prev
    Next
    INDEX
    Explanations

    code initialization and structure markers

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

    The neuron is primarily triggered by numeric tokens—any standalone numbers or version‐style digits in the code.

    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
     their
    -1.16
    의
    -1.14
     from
    -1.04
     pasien
    -1.03
    thodoxy
    -1.02
    ).—
    -1.02
     more
    -1.01
    }')
    -0.98
     even
    -0.96
    ']),
    -0.94
    POSITIVE LOGITS
    んだから
    1.31
     اوم
    1.08
    Mvh
    1.05
     можешь
    1.05
     ребё
    1.04
    According
    1.03
    خخ
    1.01
    かったです
    1.01
    Equipo
    1.00
    loroethane
    0.99
    Activations Density 0.000%

    No Known Activations