© 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. 19855
    Prev
    Next
    INDEX
    Explanations

    Code or data snippets

    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
    kie
    -0.28
    orphic
    -0.27
    çī©ä»¶
    -0.26
    athy
    -0.26
    hoc
    -0.25
     ãĥĪãĥ©ãĥĥãĤ¯
    -0.25
    .setDefault
    -0.25
    ppy
    -0.24
     metic
    -0.24
    olis
    -0.24
    POSITIVE LOGITS
    s
    0.40
    ERING
    0.31
    说
    0.29
    suit
    0.29
    OUGH
    0.28
    人
    0.28
    er
    0.28
    amil
    0.27
    åĩºåħ¥
    0.27
    degree
    0.27
    Activations Density 0.987%

    No Known Activations