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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 19066
    Prev
    Next
    INDEX
    Explanations

    some

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ä¸¥æł¼çļĦ
    -0.33
    types
    -0.31
     anything
    -0.29
    ä»»ä½ķæŶåĢĻ
    -0.29
    regions
    -0.29
    _TYPES
    -0.28
    (types
    -0.28
     groups
    -0.28
     anywhere
    -0.28
    categories
    -0.27
    POSITIVE LOGITS
     kind
    0.46
     sort
    0.41
     type
    0.40
     way
    0.40
    way
    0.37
     notion
    0.36
    kind
    0.36
    _kind
    0.35
     idea
    0.34
     form
    0.34
    Activations Density 0.918%

    No Known Activations