© 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. 129745
    Prev
    Next
    INDEX
    Explanations

    Cause and effect

    np_max-act · gemini-2.0-flash
    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
    hood
    -0.29
    ä¿Ŀåį«
    -0.29
     repetition
    -0.27
    gom
    -0.27
    带
    -0.25
     Homeland
    -0.25
    为ä¸Ńå¿ĥ
    -0.25
    -pad
    -0.25
     hood
    -0.24
     precedence
    -0.24
    POSITIVE LOGITS
    设置äºĨ
    0.27
    .setResult
    0.27
    JsonValue
    0.27
     drafted
    0.27
    EFR
    0.27
    çŃ¾ç½²äºĨ
    0.26
     setStatus
    0.26
     Draft
    0.25
     draft
    0.25
    çŃ¾ç½²
    0.25
    Activations Density 41.340%

    No Known Activations