© 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. 107986
    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
    ç³Ĭ
    -0.28
    antine
    -0.28
    èĭ¦
    -0.27
     Aub
    -0.26
     mortal
    -0.25
     apl
    -0.25
    elog
    -0.25
    éĢĨè¡Į
    -0.24
    \param
    -0.24
    çĶŁæ¶¯
    -0.24
    POSITIVE LOGITS
    llib
    0.27
     Decompiled
    0.26
    JUST
    0.26
    CLR
    0.25
    dsn
    0.25
    лиÑĨ
    0.24
    agog
    0.24
     patiently
    0.24
    .all
    0.24
    spell
    0.24
    Activations Density 0.659%

    No Known Activations

    This feature has no known activations.