© 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. 54725
    Prev
    Next
    INDEX
    Explanations

    Technical/scientific language

    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
    hir
    -0.31
    åĩłæŃ¥
    -0.30
    è¤Ĵ
    -0.27
    ä¸ĢåĽ¢
    -0.24
     dimension
    -0.24
    ndo
    -0.24
    大éĺŁ
    -0.24
     fortunate
    -0.24
    NS
    -0.24
    åĩłæ¬¡
    -0.23
    POSITIVE LOGITS
    ovi
    0.29
     tote
    0.28
    _RET
    0.28
    建çŃijçī©
    0.26
     pron
    0.26
     alive
    0.25
     building
    0.25
     recept
    0.25
     organizational
    0.24
     fellows
    0.24
    Activations Density 0.006%

    No Known Activations