© 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. 15747
    Prev
    Next
    INDEX
    Explanations

    questions about alternatives

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

    The neuron is primarily detecting numeric tokens (literal numbers, including multi-digit and decimal values).

    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
     such
    -0.90
     {?>
    -0.84
    bc
    -0.83
    hull
    -0.83
     другие
    -0.80
    cashire
    -0.79
    week
    -0.79
     they
    -0.79
     другое
    -0.78
    Ŭ
    -0.76
    POSITIVE LOGITS
     latter
    1.81
    后者
    1.45
    latter
    1.36
     Latter
    1.31
     whichever
    1.31
    どちら
    1.20
    どっち
    1.11
     optar
    1.10
     choose
    1.02
    いずれ
    1.02
    Activations Density 0.120%

    No Known Activations