© 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. 19782
    Prev
    Next
    INDEX
    Explanations

    Don

    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
    -da
    -0.27
    ТЬ
    -0.26
    apper
    -0.26
    chure
    -0.26
    dle
    -0.25
     lul
    -0.25
    تØŃد
    -0.24
    ularity
    -0.24
    bbing
    -0.24
    insula
    -0.24
    POSITIVE LOGITS
    ë¡ł
    0.32
    allen
    0.30
    icut
    0.29
     Club
    0.27
    Rom
    0.27
    isque
    0.27
     ValidationResult
    0.26
    央
    0.26
    fällt
    0.25
    ä¼ģä¸ļåľ¨
    0.25
    Activations Density 1.080%

    No Known Activations