© 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. 33-TRANSCODER-HP
    7. 52596
    Prev
    Next
    INDEX
    Explanations

    technical or code related texts

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    inals
    -0.35
    è·Į
    -0.29
     heights
    -0.29
    stown
    -0.28
    unci
    -0.26
    czÄħ
    -0.26
    bose
    -0.26
    >(()
    -0.25
     Reds
    -0.25
     coc
    -0.25
    POSITIVE LOGITS
    åľ°éģĵ
    0.27
    endregion
    0.25
    thon
    0.25
     Aure
    0.24
     kosten
    0.24
     perception
    0.24
     اÙĦطب
    0.24
    çĢ¹
    0.23
    qual
    0.23
     guid
    0.23
    Activations Density 0.012%

    No Known Activations