© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 18-TRANSCODER-HP
    7. 88415
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    缸åIJĮ
    -0.32
     identical
    -0.32
     similar
    -0.31
    éĴ§
    -0.29
    åIJĮæł·çļĦ
    -0.29
    æĹĭ
    -0.28
    ãģ¨åIJĮãģĺ
    -0.27
    çĽ¸ä¼¼
    -0.26
     sub
    -0.26
     same
    -0.26
    POSITIVE LOGITS
    濯
    0.30
    çĩĥçĥ§
    0.30
     sack
    0.28
    ropp
    0.27
    ÑĮÑĤе
    0.25
     Rehab
    0.25
    çĩĥ
    0.25
    heits
    0.24
     heights
    0.24
    è°·
    0.24
    Activations Density 9.893%

    No Known Activations