© 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. 32-TRANSCODER-HP
    7. 43245
    Prev
    Next
    INDEX
    Explanations

    non-english or code

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    wh
    -0.53
     wh
    -0.50
     Wh
    -0.48
    ä½ķ
    -0.44
    Wh
    -0.41
     whe
    -0.39
    _wh
    -0.35
    WH
    -0.34
    .wh
    -0.33
    -wh
    -0.33
    POSITIVE LOGITS
    oldt
    0.27
    [keys
    0.27
    allee
    0.26
    inee
    0.25
    PRI
    0.24
     cuid
    0.24
    elt
    0.24
    ä¸īåĽĽ
    0.23
     Gef
    0.23
    è§ģè¿ĩ
    0.23
    Activations Density 0.003%

    No Known Activations