© 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. 22-TRANSCODER-HP
    7. 130788
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åįļè§Ī
    -0.27
    ä¸Ģ件äºĭ
    -0.26
     Notre
    -0.25
    大ä¼ļ
    -0.24
    basic
    -0.24
    _basic
    -0.24
     Corpor
    -0.24
    空åĨĽ
    -0.24
    讼
    -0.24
    Cc
    -0.23
    POSITIVE LOGITS
    å¹ķ
    0.26
     marching
    0.25
    -pound
    0.25
     occup
    0.24
    ÑģÑĤÑĥпил
    0.24
    <small
    0.24
    roz
    0.23
    @student
    0.23
    ä¹Ĺ
    0.23
    @s
    0.23
    Activations Density 0.020%

    No Known Activations

    This feature has no known activations.