© 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. 23-TRANSCODER-HP
    7. 23961
    Prev
    Next
    INDEX
    Explanations

    uses

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_23.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.23.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    å¼ĢæĶ¾
    -0.27
    æıIJåįĩ
    -0.27
     ttk
    -0.27
     squeezing
    -0.26
    å·¥ä½ľä½ľé£İ
    -0.24
     squeezed
    -0.24
    æī¼
    -0.24
    _Osc
    -0.24
    nk
    -0.23
    &q
    -0.23
    POSITIVE LOGITS
     Hire
    0.31
    irth
    0.31
    梯
    0.27
     BEFORE
    0.27
    trand
    0.27
     seemingly
    0.26
    å¯ĮåĬĽ
    0.25
    epy
    0.25
     tradition
    0.25
    _BEFORE
    0.24
    Activations Density 0.004%

    No Known Activations