© 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. 31-TRANSCODER-HP
    7. 18549
    Prev
    Next
    INDEX
    Explanations

    way

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çļĦä¸Ģ
    -0.29
    º«
    -0.27
    è²Į
    -0.26
    çļĦè¯Ŀ
    -0.25
    etched
    -0.25
     nag
    -0.24
    >>>(
    -0.24
    åľ¨åĨħçļĦ
    -0.24
    Valid
    -0.24
    æľ¬çĿĢ
    -0.24
    POSITIVE LOGITS
     for
    0.38
    .for
    0.32
    for
    0.32
     For
    0.31
    _for
    0.30
    iÄĩ
    0.28
    -for
    0.27
     длÑı
    0.27
    For
    0.27
    izzare
    0.26
    Activations Density 0.070%

    No Known Activations