© 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. 94932
    Prev
    Next
    INDEX
    Explanations

    Code and citations

    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
    ç»Ī端
    -0.28
    itas
    -0.28
    /tasks
    -0.27
    èŃ¦ç¤º
    -0.27
    urv
    -0.27
     reminders
    -0.24
     lÆ°u
    -0.24
    stm
    -0.24
    æµģæ·Į
    -0.24
     green
    -0.23
    POSITIVE LOGITS
    neys
    0.30
    æĬ½å¥ĸ
    0.27
    pell
    0.27
    .Observer
    0.27
    åİŁåĪĻ
    0.26
    å½¼å¾Ĺ
    0.26
    çĸ²
    0.25
     decl
    0.25
    pixel
    0.25
     Observer
    0.25
    Activations Density 0.002%

    No Known Activations