© 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. 107354
    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
    ä¸Ģéĥ¨åĪĨ
    -0.26
     hoÅŁ
    -0.26
    GST
    -0.26
    JV
    -0.26
    主æµģ
    -0.25
     rodz
    -0.24
    ãĥ¼ãĥī
    -0.24
    Bear
    -0.23
    éĹ²
    -0.23
     JT
    -0.23
    POSITIVE LOGITS
    rett
    0.30
    å±Ĥ
    0.29
    layer
    0.28
    æį®
    0.26
     layer
    0.25
    (Un
    0.25
    ighest
    0.25
    vern
    0.24
    -layer
    0.24
    (unique
    0.24
    Activations Density 2.312%

    No Known Activations

    This feature has no known activations.