© 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. 25-TRANSCODER-HP
    7. 109514
    Prev
    Next
    INDEX
    Explanations

    horrors

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    RAND
    -0.28
    éļĶ
    -0.28
    isin
    -0.26
    /cop
    -0.26
    GTK
    -0.25
    个交æĺĵ
    -0.25
    asil
    -0.24
     acne
    -0.24
    åĮĸèĤ¥
    -0.24
     insulin
    -0.23
    POSITIVE LOGITS
    éĥ¨ç½²
    0.28
     minion
    0.27
    ataka
    0.27
    è¿Ļä¹Īå¤ļå¹´
    0.26
    å¸ĥç½®
    0.26
    带走
    0.25
    æµģåIJij
    0.25
    ç¼ľ
    0.25
    hurst
    0.25
     deployments
    0.24
    Activations Density 0.915%

    No Known Activations