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. 1-GEMMASCOPE-RES-16K
    6. 9260
    Prev
    Next
    INDEX
    Explanations

    abbreviated terms or codes that start with "PR"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.1.hook_resid_post
    Hook Layer
    1
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    بوابة
    -0.72
    Билгалдахарш
    -0.57
    ipient
    -0.53
    IGraphics
    -0.53
    gefü
    -0.53
     Elliott
    -0.53
    みな
    -0.52
    Демографія
    -0.51
    Portail
    -0.51
    ellschaft
    -0.51
    POSITIVE LOGITS
    PR
    3.28
     PR
    3.14
    pr
    3.11
     pr
    2.95
    Pr
    2.66
     Pr
    2.65
    pri
    1.63
    prn
    1.58
     Pri
    1.54
    Pri
    1.49
    Activations Density 0.085%

    No Known Activations