© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15188
    Prev
    Next
    INDEX
    Explanations

    apostrophes in the text

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

    Apostrophes preceding abbreviated words (e.g., 'cause, 'til)

    np_acts-logits-general · gemini-2.0-flash

    'cause or 'til

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    PreExecute
    -0.59
    utilisons
    -0.59
    ughty
    -0.58
    хьтан
    -0.57
     Frankel
    -0.57
    っきり
    -0.57
    ROIT
    -0.56
     Barbier
    -0.55
     Gallen
    -0.55
    tingu
    -0.53
    POSITIVE LOGITS
    cause
    0.64
    nuff
    0.57
    til
    0.56
     "'
    0.56
    nother
    0.56
     ery
    0.52
     '`
    0.52
    tis
    0.52
    ArrowToggle
    0.51
    ਚ
    0.50
    Activations Density 0.146%

    No Known Activations