© 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. 19700
    Prev
    Next
    INDEX
    Explanations

    Code and technical documentation

    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
    TON
    -0.28
    çĪµ
    -0.28
    èķī
    -0.27
    Guid
    -0.27
    baugh
    -0.26
     Guid
    -0.26
     convention
    -0.26
    íĸ¥
    -0.26
    keit
    -0.26
    éħ®
    -0.26
    POSITIVE LOGITS
    XD
    0.28
     acts
    0.27
    使åij½
    0.26
    ACT
    0.26
    ocr
    0.26
    UGHT
    0.26
    act
    0.26
    ä¸įæĪIJ
    0.26
    xr
    0.26
    .Act
    0.25
    Activations Density 1.035%

    No Known Activations