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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 7229
    Prev
    Next
    INDEX
    Explanations

    phrases related to gameplay strategies in sports.

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

    mentions of soccer (football) match events and player actions.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    articles and determiners like "the", especially when they precede nouns or start sentences. The neuron seems to activate strongly for "The" at the beginning of sentences, and shows consistent moderate activation for "the" throughout the text samples.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot

    Follows "the" or a capital letter

    np_acts-logits-general · gemini-2.0-flash

    specific words after common tokens

    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
    
    -0.48
    orte
    -0.44
    ordi
    -0.43
    orra
    -0.42
    Mitte
    -0.41
    mdi
    -0.41
    poti
    -0.41
     voulez
    -0.40
    orno
    -0.40
    ленность
    -0.40
    POSITIVE LOGITS
    MemoryWarning
    0.78
     ProtoMessage
    0.77
     المعيارى
    0.76
     visitors
    0.73
    GEBURTSDATUM
    0.72
    RTEE
    0.72
    Personensuche
    0.71
     كومونز
    0.71
    بوابة
    0.70
    ביוגרפיה
    0.70
    Activations Density 0.121%

    No Known Activations