© 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. 31-TRANSCODER-HP
    7. 19713
    Prev
    Next
    INDEX
    Explanations

    Scientific/technical articles

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    (Dense
    -0.27
    adol
    -0.26
     Arch
    -0.26
    é¼·
    -0.26
     spur
    -0.26
    è·
    -0.25
    çªģ
    -0.25
     refl
    -0.25
    forc
    -0.25
     jac
    -0.24
    POSITIVE LOGITS
     poll
    0.28
    Collapsed
    0.26
    itution
    0.26
    uts
    0.25
     tá»ķ
    0.25
     crops
    0.25
    å©¿
    0.25
    uto
    0.25
    éĶĢ
    0.24
     proof
    0.24
    Activations Density 0.148%

    No Known Activations