© 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. 130487
    Prev
    Next
    INDEX
    Explanations

    Negative outcomes/causes

    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
    Ļ¨
    -0.28
    iska
    -0.26
    ends
    -0.26
     temps
    -0.25
    èĻ±
    -0.25
    ãģĭãĤĭ
    -0.25
     mods
    -0.25
     instruments
    -0.24
    -hop
    -0.24
    ç«Ń
    -0.24
    POSITIVE LOGITS
    alytics
    0.25
    letic
    0.24
     Brief
    0.24
     Outline
    0.23
     place
    0.23
    å¤įä»ĩ
    0.23
    emic
    0.23
    @d
    0.23
    .!
    0.23
     hel
    0.23
    Activations Density 19.051%

    No Known Activations