© 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. 22-GEMMASCOPE-RES-131K
    7. 71072
    Prev
    Next
    INDEX
    Explanations

    shuffled word

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

    The neuron is picking out code tokens—i.e. function and variable names or other identifiers in programming source.

    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
    ༝
    -1.02
     oeufs
    -0.98
     comprando
    -0.96
     fenêtre
    -0.94
    dáms
    -0.91
    dragons
    -0.90
     SLIP
    -0.90
     autorizado
    -0.89
     colorir
    -0.88
    serves
    -0.88
    POSITIVE LOGITS
     ``
    1.09
     „
    1.03
    ㆍ
    1.01
     `
    0.99
     seconda
    0.98
    </h3>
    0.91
    </h5>
    0.91
    sta
    0.90
     –
    0.87
     tutti
    0.86
    Activations Density 0.090%

    No Known Activations