© 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. 96939
    Prev
    Next
    INDEX
    Explanations

    at

    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
    _equal
    -0.28
    ths
    -0.26
    _nil
    -0.25
    å®ĩå®Ļ
    -0.25
    åIJİåĨį
    -0.25
     conj
    -0.25
    .until
    -0.25
    缸
    -0.24
    wjgl
    -0.24
    æĶ¶åıĸ
    -0.24
    POSITIVE LOGITS
    åħ¶ä»ĸ
    0.36
     other
    0.36
    other
    0.31
    æŃ£ä¹ī
    0.30
     otros
    0.30
    åħ¶ä»ĸçļĦ
    0.29
    åħ¶å®ĥ
    0.28
     demás
    0.27
    compet
    0.27
     дÑĢÑĥгиÑħ
    0.27
    Activations Density 0.883%

    No Known Activations