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-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 117267
    Prev
    Next
    INDEX
    Explanations

    references to specific types of food or candy, particularly those associated with unique names or descriptors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     impulsive
    -0.33
    менно
    -0.31
     plenamente
    -0.29
     Simply
    -0.28
     most
    -0.28
    directives
    -0.27
     simply
    -0.27
    gesteld
    -0.27
     fondamentali
    -0.27
    事を
    -0.26
    POSITIVE LOGITS
    Autoritní
    0.77
    LookAnd
    0.75
    RTEX
    0.75
    featureID
    0.73
     lenker
    0.71
    webElementXpaths
    0.70
    RTEE
    0.69
     препратки
    0.66
     informée
    0.65
    DockStyle
    0.65
    Activations Density 0.038%

    No Known Activations