© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 24-GEMMASCOPE-TRANSCODER-16K
    7. 476
    Prev
    Next
    INDEX
    Explanations

    names of cities in France, especially Paris, and also dates

    oai_token-act-pair · gemini-2.0-flash

    French language and locations

    np_max-act-logits · gemini-2.0-flash

    say "place names"

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_24/width_16k/average_l0_37
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     French
    -1.01
     France
    -0.92
    French
    -0.83
     french
    -0.82
     Franse
    -0.81
     Ligue
    -0.80
     Jean
    -0.79
     francesa
    -0.78
     Paris
    -0.76
    France
    -0.76
    POSITIVE LOGITS
     Shakspeare
    0.63
     Shaksp
    0.62
     Diſ
    0.61
     Theſe
    0.60
     Reſ
    0.59
     ſeveral
    0.58
    extré
    0.57
     Jefus
    0.57
     ainfi
    0.56
     myſelf
    0.56
    Activations Density 0.766%

    No Known Activations