© 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. 13468
    Prev
    Next
    INDEX
    Explanations

    RUffIn

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

    The neuron is essentially measuring how many tokens occur between the <start> and <end> markers—that is, it activates in proportion to the length of each text block.

    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
     Voraussetzung
    -1.12
     meub
    -1.09
    Already
    -1.09
     norvég
    -1.04
     Lég
    -1.03
     HelloWorld
    -1.03
     seront
    -1.02
     MICHAEL
    -1.02
     Introducción
    -1.02
     deklar
    -1.01
    POSITIVE LOGITS
     has
    1.23
     and
    1.18
     on
    1.17
     after
    1.07
     provides
    1.02
    eynman
    1.00
     before
    0.98
    1
    0.98
     rudimentary
    0.98
     it
    0.96
    Activations Density 0.001%

    No Known Activations