© 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. 95601
    Prev
    Next
    INDEX
    Explanations

    blog posts

    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
    .remaining
    -0.27
     numbering
    -0.25
    丢äºĨ
    -0.25
    enerative
    -0.25
    ÑĩеÑģÑĤво
    -0.24
    _Copy
    -0.24
     Remaining
    -0.24
    wife
    -0.24
    urities
    -0.24
    urai
    -0.24
    POSITIVE LOGITS
    æĹ¶ä»£
    0.30
     breakthrough
    0.29
    fo
    0.28
    æݨ
    0.28
    èĢĥèĻij
    0.28
     feas
    0.27
    è¿ij
    0.27
    æħ¨
    0.27
    еÑĤÑģÑı
    0.26
    æĿ¾
    0.26
    Activations Density 0.105%

    No Known Activations