© 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. 54726
    Prev
    Next
    INDEX
    Explanations

    code and software

    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
     agency
    -0.36
     Agency
    -0.32
    agency
    -0.31
    Agency
    -0.27
     agent
    -0.27
    gd
    -0.27
    lt
    -0.27
    Agent
    -0.26
    奥æĸ¯
    -0.26
    lay
    -0.26
    POSITIVE LOGITS
     rail
    0.25
    æħ¢
    0.24
    modulo
    0.24
     couples
    0.24
     Coupons
    0.23
    åŁºå»º
    0.23
     Rail
    0.23
    å½ķåĪ¶
    0.23
    idon
    0.23
    .Arg
    0.23
    Activations Density 0.005%

    No Known Activations