© 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. 41919
    Prev
    Next
    INDEX
    Explanations

    Arguments and publications

    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
    ancies
    -0.27
    åĩĨå¤ĩå·¥ä½ľ
    -0.27
    æ³Ľ
    -0.27
    åºĹ
    -0.26
    arend
    -0.25
    omencl
    -0.25
    åĻ¨
    -0.25
    æIJŃ
    -0.24
     pan
    -0.24
    è¯ģ
    -0.24
    POSITIVE LOGITS
    edi
    0.31
     probes
    0.29
     sounding
    0.28
     Probe
    0.28
    reflect
    0.28
     probing
    0.28
     probe
    0.27
     correctly
    0.27
     Cont
    0.27
    -cont
    0.27
    Activations Density 0.002%

    No Known Activations