© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 16k
    9. 18-GEMMASCOPE-RES-16K
    11. 14298
    Prev
    Next
    INDEX
    Explanations

    technical or specialized terms related to genes and genetic analysis

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

    preceding specific measurement units

    np_acts-logits-general · gemini-2.0-flash

    words in the real world

    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
    ########.
    -0.79
    tagHelperRunner
    -0.72
    +#+#
    -0.70
    ScopeManager
    -0.68
    ंदीखरीदारी
    -0.67
    saraba
    -0.61
    #+#
    -0.61
    AddHtmlAttribute
    -0.60
    }{*}{
    -0.59
    __()
    -0.59
    POSITIVE LOGITS
     but
    0.77
     namun
    0.60
     antaranya
    0.59
     namely
    0.59
    而是
    0.57
     tetapi
    0.57
     sondern
    0.56
     Например
    0.56
     habituellement
    0.54
    but
    0.52
    Activations Density 1.096%

    No Known Activations