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 - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 926
    Prev
    Next
    INDEX
    Explanations

    API-related actions and HTTP request methods

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

    network-related functions and methods for fetching data, specifically in the context of JavaScript programming

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

    http requests and data retrieval

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     pinulongan
    -0.88
     Monfieur
    -0.78
    "]);
    
    -0.77
     myſelf
    -0.77
     Efq
    -0.77
    ſelf
    -0.76
    ValueStyle
    -0.72
    XmlAccessType
    -0.72
     Anſ
    -0.71
    >");
    
    -0.71
    POSITIVE LOGITS
     https
    0.55
     http
    0.50
    WithEmail
    0.48
    encodeWith
    0.48
    makeConstraints
    0.46
    tahui
    0.45
    BASEPATH
    0.43
    api
    0.42
    http
    0.42
     et
    0.41
    Activations Density 0.012%

    No Known Activations