© 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. 20-TRANSCODER-HP
    7. 64750
    Prev
    Next
    INDEX
    Explanations

    maximum

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    atches
    -0.27
    ç´ł
    -0.27
    ç¼
    -0.27
    ç´łè´¨æķĻèĤ²
    -0.25
     conform
    -0.25
    å®ŀéĻħæĥħåĨµ
    -0.25
    æĭ¨
    -0.24
     DEFIN
    -0.24
    ote
    -0.24
    æ½
    -0.24
    POSITIVE LOGITS
    ëłī
    0.28
    cod
    0.26
     Behavior
    0.25
    jax
    0.25
    rä
    0.24
    loi
    0.24
    avored
    0.24
    以ä¸ĬçļĦ
    0.24
    morph
    0.24
    ÑĥпиÑĤÑĮ
    0.24
    Activations Density 4.391%

    No Known Activations