© 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. 35-TRANSCODER-HP
    7. 38808
    Prev
    Next
    INDEX
    Explanations

    Software licenses

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    çĶ¨æĪ·çļĦ
    -0.31
    dens
    -0.30
    defines
    -0.28
    è¿Ļæł·çļĦ
    -0.28
    atus
    -0.28
    Will
    -0.27
    will
    -0.26
    much
    -0.26
    Exp
    -0.25
    disk
    -0.25
    POSITIVE LOGITS
     neighbor
    0.28
    å°±æĪIJäºĨ
    0.27
    社
    0.27
    åIJİèĢħ
    0.27
     both
    0.25
     together
    0.25
    年第
    0.25
    çļĦçłĶåıij
    0.25
     neighbors
    0.24
    é¢ij
    0.24
    Activations Density 0.029%

    No Known Activations