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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1797
    Prev
    Next
    INDEX
    Explanations

    phrases related to instructions or operational processes

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

    Followed by "væ" or "по"

    np_acts-logits-general · gemini-2.0-flash

    Danish labor movements

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    følge
    -0.67
    CloseOperation
    -0.60
    usermodel
    -0.57
    zulegen
    -0.56
    voorbeeld
    -0.55
    zonder
    -0.55
    foglal
    -0.55
    достатки
    -0.54
     BoxDecoration
    -0.54
     kasarigan
    -0.53
    POSITIVE LOGITS
    take
    0.59
    по
    0.58
    hang
    0.57
    sp
    0.57
    wind
    0.56
    print
    0.55
    âm
    0.54
     pulumi
    0.54
    กัด
    0.53
    going
    0.52
    Activations Density 0.044%

    No Known Activations