© 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. 130590
    Prev
    Next
    INDEX
    Explanations

    This neuron fires most strongly on sentence- or clause-initial tokens—e.g. the first word of a new sentence or narrative beat.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    followed by capitalized words

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     Result
    -1.00
    Adds
    -0.96
    IsIn
    -0.93
     ziyaret
    -0.90
     those
    -0.90
     Gewebe
    -0.90
     conejo
    -0.89
     Adds
    -0.87
     creates
    -0.87
     desain
    -0.86
    POSITIVE LOGITS
    immediately
    1.02
    centerTitle
    1.01
     proceeded
    0.99
     immediately
    0.98
    ActionButton
    0.90
    立即
    0.90
     lacked
    0.86
     Immediately
    0.86
    }}^{\
    0.84
     дальней
    0.84
    Activations Density 0.061%

    No Known Activations