© 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-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 15387
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and spaces

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

    Numbers written as words

    np_acts-logits-general · gemini-2.0-flash

    numbers and quantities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     fué
    -0.73
     فريبيس
    -0.70
     europé
    -0.63
     illustrationer
    -0.63
     estimés
    -0.62
    FontOfSize
    -0.61
    ान्त
    -0.61
    parsedMessage
    -0.61
    withIdentifier
    -0.61
    unjukan
    -0.60
    POSITIVE LOGITS
     Reverso
    0.58
     eight
    0.55
     several
    0.55
     six
    0.53
     nearly
    0.53
    timedelta
    0.52
    ten
    0.51
     seven
    0.51
    InputBorder
    0.50
     nine
    0.49
    Activations Density 0.270%

    No Known Activations