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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 3170
    Prev
    Next
    INDEX
    Explanations

    terms related to classification or numerical sequences

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

    an observed direction or pattern

    np_acts-logits-general · gemini-2.0-flash

    trend and pattern

    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
     imagined
    -0.55
     supposes
    -0.52
    enumi
    -0.51
     myſelf
    -0.51
    ikirkan
    -0.51
     henkilö
    -0.51
    udessa
    -0.50
     scalars
    -0.50
     sauvage
    -0.50
     imagines
    -0.49
    POSITIVE LOGITS
     trend
    2.48
     trends
    2.32
     pattern
    2.20
    trend
    2.09
     patterns
    2.00
     Trend
    1.96
     Trends
    1.88
    trends
    1.87
    pattern
    1.87
     Pattern
    1.85
    Activations Density 0.680%

    No Known Activations