© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 27555
    Prev
    Next
    INDEX
    Explanations

    expressions of encouragement and invitations to take action

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    hesitate to, free to, afraid to

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     surla
    -0.71
    Espèce
    -0.63
     queſta
    -0.61
     Menſchen
    -0.60
    iſen
    -0.60
    ágenes
    -0.60
    <end_of_turn>
    -0.60
    pantalones
    -0.59
    <unused27>
    -0.58
     erſten
    -0.58
    POSITIVE LOGITS
     please
    0.37
     to
    0.36
    참고
    0.33
    please
    0.31
     Please
    0.30
     welcome
    0.30
     SM
    0.29
     pot
    0.29
    SM
    0.29
     bienven
    0.28
    Activations Density 0.002%

    No Known Activations