© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 9-GEMMASCOPE-RES-131K
    11. 102455
    Prev
    Next
    INDEX
    Explanations

    references to work-related experiences or interactions

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot

    talk about foreign languages or statements

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

    short emphatic discourse markers and interjections (phrases like "Talk about", "Typical", "Turns out", "No way", "Believe it or not") that signal commentary or emphasis.

    oai_token-act-pair · gpt-5-mini
    Triggered by @vetterc0

    the neuron detects discourse markers and short comment/transition phrases that signal emphasis, reactions, or narrative asides (e.g., "turns out", "despite the fact", "typical", "talk about").

    oai_token-act-pair · gpt-5-mini
    Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
     NSCoder
    -0.67
    rungsseite
    -0.64
    RTSC
    -0.60
    Rup
    -0.59
    imet
    -0.57
    ivant
    -0.57
     Rup
    -0.57
    hland
    -0.57
     disambiguazione
    -0.57
     Rescue
    -0.56
    POSITIVE LOGITS
     observaciones
    0.43
     päivä
    0.40
     oiseaux
    0.39
     acuer
    0.39
    WithIOException
    0.39
     hongos
    0.39
     ensayo
    0.39
     conejo
    0.39
     mariposas
    0.39
     creencias
    0.39
    Activations Density 0.335%

    No Known Activations