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. 13864
    Prev
    Next
    INDEX
    Explanations

    specific alphanumeric symbols and structured code syntax

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

    Capital letter B, often followed by another letter

    np_acts-logits-general · gemini-2.0-flash

    ball grid array or bacillus

    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
    twimg
    -0.48
     سكانية
    -0.44
     VizieR
    -0.44
    ṳ
    -0.41
    couz
    -0.41
     poulet
    -0.41
    ganet
    -0.41
    glTexCoord
    -0.40
    ríguez
    -0.40
     Atiku
    -0.38
    POSITIVE LOGITS
     B
    0.97
    B
    0.89
     b
    0.88
     BB
    0.84
    beta
    0.83
    b
    0.79
     beta
    0.78
     bio
    0.78
     Bio
    0.78
     BI
    0.78
    Activations Density 1.772%

    No Known Activations