© 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. 30-TRANSCODER-HP
    7. 96778
    Prev
    Next
    INDEX
    Explanations

    code, math

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     equ
    -0.49
     scal
    -0.34
     Equ
    -0.34
     acute
    -0.30
    Equ
    -0.28
     escal
    -0.28
    eldo
    -0.28
    equ
    -0.27
    keepers
    -0.27
     tails
    -0.27
    POSITIVE LOGITS
     convex
    0.31
    raf
    0.28
    ç¹ŀ
    0.27
    fan
    0.26
    Ñģад
    0.26
    agraph
    0.26
    _Copy
    0.26
     GPLv
    0.25
    (regex
    0.25
    cribe
    0.25
    Activations Density 0.884%

    No Known Activations