© 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. 41772
    Prev
    Next
    INDEX
    Explanations

    bel

    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
     Citation
    -0.28
    Pane
    -0.26
     gating
    -0.25
    æİĮéŨ
    -0.24
     damaging
    -0.24
     бÑĥмаг
    -0.24
    cards
    -0.24
     damage
    -0.23
    å¹¿æ³Ľ
    -0.23
    åºĦ
    -0.23
    POSITIVE LOGITS
    .bo
    0.28
    宵
    0.27
    avy
    0.27
    _configs
    0.27
     Maui
    0.27
    ittle
    0.26
    æ·±åħ¥æİ¨è¿Ľ
    0.26
    èħ»
    0.25
    沸
    0.25
    athers
    0.24
    Activations Density 0.001%

    No Known Activations