© 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. 18506
    Prev
    Next
    INDEX
    Explanations

    block

    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
    iasi
    -0.27
    elry
    -0.26
    èĬĴ
    -0.24
     Mandatory
    -0.23
    ierce
    -0.23
    jee
    -0.23
     Heaven
    -0.22
    æĹ·
    -0.22
    ÑĪе
    -0.22
    disposing
    -0.22
    POSITIVE LOGITS
     block
    1.90
    åĿĹ
    1.74
     blocks
    1.67
    block
    1.65
     Block
    1.61
    -block
    1.54
    Block
    1.53
    _block
    1.45
     Blocks
    1.44
    blocks
    1.43
    Activations Density 20.431%

    No Known Activations