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. 886
    Prev
    Next
    INDEX
    Explanations

    references to Stanford University and its involvement in discussions on affirmative action and diversity

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

    mentions of Stanford University and its associated events or decisions

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

    Stanford, HP, Silicon Valley

    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
     Tyne
    -0.58
     Juneau
    -0.52
    BIS
    -0.51
     Stok
    -0.47
    ziak
    -0.46
     Lond
    -0.46
     Brooklyn
    -0.46
     Marseille
    -0.45
     coca
    -0.45
    uxxxx
    -0.45
    POSITIVE LOGITS
    +#+#
    0.94
     Palo
    0.75
     Stanford
    0.73
     Silicon
    0.72
     createState
    0.72
    MigrationBuilder
    0.71
    Palo
    0.70
    Silicon
    0.69
    createCanvas
    0.68
    Stanford
    0.67
    Activations Density 0.082%

    No Known Activations