© 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. 10-GEMMASCOPE-RES-131K
    7. 15227
    Prev
    Next
    INDEX
    Explanations

    F20 or numbers

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

    The neuron detects percent-encoded URL escape sequences (i.e. “%” followed by two hex 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_10/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
     unrelenting
    -2.83
     burgeoning
    -2.50
     Surprisingly
    -2.48
     diligent
    -2.28
     unparalleled
    -2.22
     even
    -2.22
     after
    -2.22
     enticing
    -2.20
     when
    -2.19
     longstanding
    -2.19
    POSITIVE LOGITS
    ఽ
    2.58
    踯
    2.41
     nieuwe
    2.41
     strikingly
    2.39
     fantastic
    2.38
     insanely
    2.38
     molti
    2.36
     REALLY
    2.30
     decidedly
    2.30
     soooo
    2.28
    Activations Density 0.008%

    No Known Activations