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. 18-GEMMASCOPE-RES-16K
    6. 1562
    Prev
    Next
    INDEX
    Explanations

    mentions of organizations or institutions, specifically related to health and research contexts

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

    "Center" or "Centre"

    np_acts-logits-general · gemini-2.0-flash

    centre of attention or gravity

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    abetes
    -0.82
    UrlResolution
    -0.71
    asthan
    -0.69
    rrggbb
    -0.66
    HostException
    -0.66
    AppDelegate
    -0.65
     Codable
    -0.64
    __*/
    -0.64
    shed
    -0.63
    httphttps
    -0.63
    POSITIVE LOGITS
     gravity
    0.84
     centers
    0.66
    Cen
    0.63
    pieces
    0.62
     Cen
    0.61
     attention
    0.59
     centres
    0.59
    piece
    0.56
     gravedad
    0.56
     Centers
    0.55
    Activations Density 0.093%

    No Known Activations