© 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. 95637
    Prev
    Next
    INDEX
    Explanations

    American

    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
    inea
    -0.29
    åĴ¥
    -0.26
     Compatible
    -0.25
    release
    -0.25
    slide
    -0.25
     releasing
    -0.25
    éĹ¼
    -0.25
    èĦ±
    -0.24
    çļĦä¸Ģåijĺ
    -0.24
    uckle
    -0.24
    POSITIVE LOGITS
     losing
    0.30
    richt
    0.27
    pi
    0.27
    ä»ĺãģij
    0.27
    á»ijc
    0.27
     economy
    0.26
    ç»ij
    0.26
    çµIJ
    0.25
     loss
    0.24
    .meta
    0.24
    Activations Density 0.101%

    No Known Activations