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. 1038
    Prev
    Next
    INDEX
    Explanations

    technical terms and identifiers related to programming and computer functions

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

    references to JIT (Just-In-Time) compilation and related statuses

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

    function calls or registry operations

    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
     Exactos
    -0.45
     rien
    -0.40
     selamat
    -0.40
    cee
    -0.40
     pasto
    -0.39
     sin
    -0.37
    APPRO
    -0.37
    イプ
    -0.37
     aceptas
    -0.36
    یں
    -0.35
    POSITIVE LOGITS
    URLException
    0.93
    ArgsConstructor
    0.83
    gameserver
    0.81
     kasarigan
    0.79
    
    0.78
    EndContext
    0.78
     Wikimedijinoj
    0.77
     Italijanski
    0.76
     tartalomajánló
    0.73
    
    0.71
    Activations Density 0.043%

    No Known Activations