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

    Neuronpedia

    APICircuit 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. 20-GEMMASCOPE-RES-16K
    6. 4160
    Prev
    Next
    INDEX
    Explanations

    terms related to programming and technical specifications, particularly focusing on settings and configurations

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

    before device, models, or phones

    np_acts-logits-general · gemini-2.0-flash

    modes, states, types

    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
     betweenstory
    -1.07
     الرياضيه
    -0.92
     للمعارف
    -0.88
    Personensuche
    -0.87
    ViewFeatures
    -0.84
    tvguidetime
    -0.83
    apimachinery
    -0.81
    ConstraintMaker
    -0.76
     Theſe
    -0.76
    PerformLayout
    -0.76
    POSITIVE LOGITS
     mode
    0.99
     version
    0.88
     category
    0.74
    の場合
    0.64
     only
    0.64
    only
    0.63
     Mode
    0.62
     or
    0.61
     option
    0.59
     maupun
    0.58
    Activations Density 1.778%

    No Known Activations