© 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. 94962
    Prev
    Next
    INDEX
    Explanations

    mixed code, file paths, text

    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
    Negative Logits
     bra
    -0.29
     Kurd
    -0.24
    åħī彩
    -0.24
    å¢ŀéķ¿çİĩ
    -0.24
    esters
    -0.24
     kosher
    -0.24
    æ±ŁåĮĹ
    -0.23
    .Release
    -0.23
     tasting
    -0.23
     bay
    -0.23
    POSITIVE LOGITS
    驹
    0.34
    ron
    0.29
    åĢĴ
    0.26
    деÑĤ
    0.26
    finity
    0.26
     WON
    0.26
    åįĴ
    0.25
    è´£
    0.25
    abar
    0.25
     Dix
    0.25
    Activations Density 0.006%

    No Known Activations