© 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. 18737
    Prev
    Next
    INDEX
    Explanations

    one

    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
    ä¸ĭ乡
    -0.27
    ungle
    -0.26
    éķ¿è¿ľ
    -0.25
    à¸Ńาà¸ģาศ
    -0.25
    otten
    -0.25
    alam
    -0.24
    ulong
    -0.24
     bottleneck
    -0.23
     Hass
    -0.23
    Ø´ÙĨ
    -0.23
    POSITIVE LOGITS
     Standing
    0.26
    ħ
    0.25
    çŀ©
    0.25
    åºľ
    0.25
    san
    0.25
    æĤ¬
    0.25
    诽
    0.24
    trusted
    0.24
    town
    0.24
    hf
    0.24
    Activations Density 0.953%

    No Known Activations