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. 1202
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and notation, particularly focused on formatting and structural elements

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

    special mathematical symbols or characters in a text

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

    LaTeX math closing braces

    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
    دانشنامهٔ
    -0.80
     lenker
    -0.74
    :✨
    -0.67
    évaluateur
    -0.63
    المكان
    -0.62
     tfsi
    -0.61
    besondere
    -0.60
     cdti
    -0.59
     רבה
    -0.59
    География
    -0.59
    POSITIVE LOGITS
    mrow
    0.85
    AxisAlignment
    0.63
    else
    0.59
     else
    0.58
    Close
    0.54
     close
    0.53
    close
    0.52
     zaw
    0.52
    izo
    0.51
    μένα
    0.51
    Activations Density 0.038%

    No Known Activations