© 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. 30-TRANSCODER-HP
    7. 95985
    Prev
    Next
    INDEX
    Explanations

    engine

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    éŀĺ
    -0.31
     PROF
    -0.28
    HING
    -0.27
    çİ·
    -0.27
    inks
    -0.26
    aira
    -0.26
    åij½åIJį
    -0.26
    æĿ¾å¼Ľ
    -0.25
    ä¸ĵç§ij
    -0.25
    ĥģ
    -0.25
    POSITIVE LOGITS
     engine
    0.33
    ừng
    0.32
    engine
    0.31
     horsepower
    0.30
    èĪįå¾Ĺ
    0.29
    gaard
    0.28
    horn
    0.28
    容
    0.28
     Engine
    0.27
    çĩĥçĥ§
    0.27
    Activations Density 0.890%

    No Known Activations