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-2-2B
    4. Residual Stream - 16k
    5. 15-GEMMASCOPE-RES-16K
    6. 2710
    Prev
    Next
    INDEX
    Explanations

    programming comments and annotations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    stdc
    -0.62
    yarnpkg
    -0.57
     Daley
    -0.56
     soi
    -0.55
     chắn
    -0.55
    mung
    -0.54
    ClInclude
    -0.52
    úly
    -0.52
     sei
    -0.52
     Cher
    -0.51
    POSITIVE LOGITS
     للمعارف
    1.08
    Jeografia
    0.90
    例句
    0.86
    новниш
    0.85
    principalColumn
    0.85
     للاسماء
    0.85
    RTEX
    0.81
    انيف
    0.78
    niająca
    0.77
    Gambas
    0.76
    Activations Density 0.136%

    No Known Activations