© 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. 42002
    Prev
    Next
    INDEX
    Explanations

    This

    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
     works
    -0.33
     ÑĢабоÑĤаеÑĤ
    -0.29
     Works
    -0.29
     dominates
    -0.27
     need
    -0.27
     represents
    -0.27
     working
    -0.27
    valueOf
    -0.26
    çŃīçĿĢ
    -0.26
    need
    -0.25
    POSITIVE LOGITS
     points
    0.37
    points
    0.32
    çĤ¹
    0.31
     Points
    0.31
    ä½Ļ
    0.30
    held
    0.29
    auc
    0.29
    çŁĽçĽ¾
    0.29
    _POINTS
    0.28
     point
    0.28
    Activations Density 0.007%

    No Known Activations