© 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. 18601
    Prev
    Next
    INDEX
    Explanations

    programming contexts

    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
    ugi
    -0.26
    pin
    -0.25
     Forever
    -0.25
     Faith
    -0.25
    åĪĹ
    -0.25
     Hop
    -0.24
    Expires
    -0.24
     Leading
    -0.24
    Hop
    -0.24
    dh
    -0.23
    POSITIVE LOGITS
    æĢ»æĶ¶åħ¥
    0.29
     UNU
    0.27
    Prem
    0.27
     bcm
    0.25
     Prem
    0.25
    otted
    0.25
    .Non
    0.25
    èħ¾
    0.25
     Sussex
    0.25
    iane
    0.25
    Activations Density 0.110%

    No Known Activations