© 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. 20-TRANSCODER-HP
    7. 130445
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Vibr
    -0.29
    æīģ
    -0.27
    æĢİèĥ½
    -0.25
    æĹı
    -0.25
    addChild
    -0.24
    ÑģÑĤвенно
    -0.24
    :",↵
    -0.24
    adle
    -0.23
    дÑĥ
    -0.23
    ComputedStyle
    -0.23
    POSITIVE LOGITS
    å¯ħ
    0.29
    åįıä¼ļä¼ļåijĺ
    0.28
     rew
    0.28
    ç±į
    0.27
     killer
    0.26
    saida
    0.26
    yme
    0.25
    迹
    0.25
     killers
    0.24
    itta
    0.24
    Activations Density 0.333%

    No Known Activations

    This feature has no known activations.