Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 1403
    Prev
    Next
    INDEX
    Explanations

    the beginning of a section or a new topic in the document

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

    Text after periods

    np_acts-logits-general · gemini-2.0-flash

    mathematical and technical terms

    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
     pran
    -0.48
     Holliday
    -0.48
    complexType
    -0.48
    tschaft
    -0.47
    ventricular
    -0.47
    HasForeignKey
    -0.46
     kite
    -0.45
    疱
    -0.44
    GOTREF
    -0.43
    ثيق
    -0.42
    POSITIVE LOGITS
    gdx
    1.64
    jgl
    1.22
    Gdx
    1.06
    badlogic
    0.95
    lwjgl
    0.85
    Билгалдахарш
    0.80
     חיצוניים
    0.76
     autorytatywna
    0.73
    Lw
    0.72
     AssemblyTitle
    0.68
    Activations Density 0.015%

    No Known Activations