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-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15061
    Prev
    Next
    INDEX
    Explanations

    references to specific locations and street names

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

    appearing within street addresses

    np_acts-logits-general · gemini-2.0-flash

    street names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
     prosi
    -0.43
    хьтан
    -0.43
    出版年
    -0.43
    nextDouble
    -0.39
    برج
    -0.37
    complete
    -0.37
    initializeApp
    -0.36
     برج
    -0.36
    pg
    -0.36
     complete
    -0.35
    POSITIVE LOGITS
     street
    0.62
     Straße
    0.57
     streets
    0.56
     straat
    0.53
     STREET
    0.53
    gasse
    0.53
     Strasse
    0.52
     Street
    0.50
    Street
    0.50
    CopyWith
    0.49
    Activations Density 0.112%

    No Known Activations