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. 14424
    Prev
    Next
    INDEX
    Explanations

    certain numeric patterns and special characters in code or data entries

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

    A period followed by a package or import statement

    np_acts-logits-general · gemini-2.0-flash

    kotlin psi hibernate hql services

    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
     TextAppearance
    -0.67
     kasarigan
    -0.67
     NSCoder
    -0.61
    AlterField
    -0.59
    DoubleQuotes
    -0.59
    Bronnen
    -0.58
     Paglinawan
    -0.57
    帖最后由
    -0.56
     AssemblyCompany
    -0.54
    djangoproject
    -0.53
    POSITIVE LOGITS
     pleaſure
    0.80
     houſe
    0.75
     reaſon
    0.74
     Majefty
    0.73
     ſch
    0.72
     Jefus
    0.71
     poffible
    0.69
     ſmall
    0.68
     purpoſe
    0.68
     myſelf
    0.68
    Activations Density 0.133%

    No Known Activations