© 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. 130986
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on enthusiastic first‐person “testimonial” statements or quoted praise (e.g. “It’s awesome,” “You have to do something more to get the phone ringing,” “I think people often make the mistake…”) — in short, it detects promotional or testimonial quotes.

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

    people and community statements

    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
     keineswegs
    -1.22
     dennoch
    -1.22
     trotzdem
    -1.09
    Caracteristici
    -1.09
     copias
    -1.08
     willkommen
    -1.07
     зелё
    -1.05
    besondere
    -1.04
    Nevertheless
    -1.03
    butuhkan
    -1.03
    POSITIVE LOGITS
     people
    1.55
     what
    1.38
     we
    1.37
     I
    1.29
     really
    1.27
     everybody
    1.25
     this
    1.14
     my
    1.13
     our
    1.11
     everyone
    1.09
    Activations Density 0.005%

    No Known Activations