Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-2-2B
    3. 21-GEMMASCOPE-TRANSCODER-16K
    4. 5594
    Prev
    Next
    INDEX
    Explanations

    sentences with scientific or technical language, often about chemistry, quantum physics, or medicine

    oai_token-act-pair · gemini-2.0-flash

    words indicating novelty or rarity

    np_token-act-pair-logits · gpt-4o-mini

    previously unknown

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.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
    ThroughAttribute
    -0.71
     protoimpl
    -0.70
    PhysRevD
    -0.67
    AnimationsModule
    -0.66
    ArrowToggle
    -0.63
    AndEndTag
    -0.61
    XtraGrid
    -0.61
    |}{$
    -0.60
     PARSER
    -0.59
    apimachinery
    -0.57
    POSITIVE LOGITS
     previously
    2.00
     hitherto
    1.64
    previously
    1.61
     heretofore
    1.58
     never
    1.58
     Previously
    1.57
     rarely
    1.43
    Previously
    1.41
     ранее
    1.39
     unknown
    1.34
    Activations Density 3.877%

    No Known Activations