© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 73947
    Prev
    Next
    INDEX
    Explanations

    proper nouns and dates

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

    This neuron responds to numeric tokens, i.e. it activates on numbers and numeric literals (e.g. constants, diff line counts, timestamps) in code.

    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
     parfü
    -1.30
    recen
    -1.25
     betek
    -1.19
     parfüm
    -1.15
     quinze
    -1.14
     Karakter
    -1.14
    érienne
    -1.13
     tribus
    -1.13
     burbu
    -1.12
     antik
    -1.11
    POSITIVE LOGITS
    Z
    1.20
    +
    1.16
    It
    1.15
    ?
    1.13
     and
    1.11
     
    1.10
    b
    1.07
    "
    1.05
    tt
    1.05
     from
    1.05
    Activations Density 0.035%

    No Known Activations