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. 22-GEMMASCOPE-TRANSCODER-16K
    4. 14440
    Prev
    Next
    INDEX
    Explanations

    pronouns and possessive pronouns with a specific focus on possessive association with "hand" and descriptions of physical interactions, especially those of a violent or sexual nature.

    oai_token-act-pair · gemini-2.0-flash

    Body parts and possessives

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
    DeleteMapping
    -0.56
     partielle
    -0.52
    esterno
    -0.51
    ModelState
    -0.47
     forskning
    -0.47
    fantasia
    -0.47
    GenerationType
    -0.47
     alem
    -0.46
    BorderLayout
    -0.44
    ufnahme
    -0.44
    POSITIVE LOGITS
    ANTLR
    0.73
    TintMode
    0.66
     AssemblyProduct
    0.66
     ivelany
    0.66
     EnglishChoose
    0.63
    titleMargin
    0.63
     noten
    0.63
    HtmlAttribute
    0.63
    ArrowToggle
    0.62
    abestanden
    0.60
    Activations Density 1.596%

    No Known Activations