© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 74888
    Prev
    Next
    INDEX
    Explanations

    percent followed by to or in

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

    The neuron selectively activates on numeric tokens—especially percentages and other statistical figures.

    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
    日発売
    -1.19
    هرة
    -1.18
    價格
    -1.18
    Under
    -1.15
     which
    -1.13
     deeply
    -1.09
     begleitet
    -1.09
     cremallera
    -1.07
    álise
    -1.07
    価格
    -1.07
    POSITIVE LOGITS
     in
    2.36
     to
    1.63
     last
    1.59
     compared
    1.49
     according
    1.29
     besonder
    1.26
    despite
    1.25
    囮
    1.24
     after
    1.23
    compared
    1.23
    Activations Density 0.010%

    No Known Activations