© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 64562
    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
    piar
    -0.28
    éĢĤ
    -0.26
    pyx
    -0.26
    enschaft
    -0.25
    èĥ³èĨĬ
    -0.25
    .har
    -0.24
    ault
    -0.24
    èĮ±
    -0.24
    Shoot
    -0.23
    adin
    -0.23
    POSITIVE LOGITS
    讽
    0.32
    è¨ĢãģĨ
    0.27
    ringe
    0.26
    otch
    0.25
    -channel
    0.25
    +":
    0.25
    å±ķå¼Ģ
    0.25
    £¼
    0.24
     channel
    0.24
    æ¸ł
    0.24
    Activations Density 6.155%

    No Known Activations

    This feature has no known activations.