© 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. 27688
    Prev
    Next
    INDEX
    Explanations

    referring to specific items or types above others

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

    The neuron fires on words that signal enumeration or elaboration in procedural or technical instructions (e.g. “—,” “preferably,” “one,” “include,” “divided,” “into,” “categories”).

    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
     and
    -1.33
     excl
    -1.06
    就连
    -1.01
     inve
    -1.00
    こういった
    -0.99
    要不是
    -0.95
     Exteriores
    -0.94
    僚
    -0.93
    similar
    -0.93
     simila
    -0.93
    POSITIVE LOGITS
     ones
    2.03
     quelli
    1.40
     those
    1.34
     ours
    1.25
     jenis
    1.20
     celui
    1.20
     from
    1.13
     īpa
    1.12
     úgy
    1.07
     niitä
    1.07
    Activations Density 0.129%

    No Known Activations