© 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. 19209
    Prev
    Next
    INDEX
    Explanations

    chemical prefixes

    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
    æĿ¥ä¸įåıĬ
    -0.28
    imens
    -0.27
    æĹ©æĻļ
    -0.26
    央
    -0.26
    fifo
    -0.26
    еÑĢж
    -0.26
    èµ°å»Ĭ
    -0.24
    èĮ¬
    -0.24
    inals
    -0.24
    aled
    -0.23
    POSITIVE LOGITS
     kata
    0.27
     sal
    0.26
     fil
    0.24
    áĥĵ
    0.24
    ilo
    0.24
    hop
    0.24
     Ded
    0.23
    dep
    0.23
     sale
    0.23
    íĸĪëĬĶëį°
    0.23
    Activations Density 0.007%

    No Known Activations