© 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. 96927
    Prev
    Next
    INDEX
    Explanations

    Journal names or code snippets

    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
     oversh
    -0.30
    ç³»ç»Ł
    -0.29
     extract
    -0.29
     systems
    -0.29
    ç³»ç»ŁçļĦ
    -0.28
     system
    -0.27
    ngr
    -0.27
    system
    -0.25
    LABEL
    -0.25
    /filter
    -0.25
    POSITIVE LOGITS
     unlocks
    0.28
    olon
    0.26
    å½¹
    0.25
    Mirror
    0.25
     unlocking
    0.25
    çī¹èī²
    0.25
    èĪ
    0.24
    commit
    0.24
    unlock
    0.24
     Commit
    0.24
    Activations Density 0.001%

    No Known Activations