© 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. 31-TRANSCODER-HP
    7. 19831
    Prev
    Next
    INDEX
    Explanations

    Est

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    è¿Ļè¾¹
    -0.30
    äºĨä¸įå°ij
    -0.30
    ãģªãģĬ
    -0.27
     plainly
    -0.27
    ission
    -0.26
     underworld
    -0.26
     Tues
    -0.26
    focused
    -0.26
    äºĨä¸ĢäºĽ
    -0.26
    essment
    -0.25
    POSITIVE LOGITS
     which
    0.38
    which
    0.34
    Which
    0.32
     Which
    0.31
     solve
    0.31
     dé
    0.29
    resolve
    0.29
    solve
    0.29
     cosa
    0.28
     cui
    0.28
    Activations Density 1.732%

    No Known Activations