© 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. 19882
    Prev
    Next
    INDEX
    Explanations

    Technical/business-related concepts

    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
    benef
    -0.32
    Benef
    -0.29
     Benef
    -0.28
    åĪ©çĽĬ
    -0.28
     benefits
    -0.28
     has
    -0.27
     béné
    -0.27
    è¿Ļä¸ĢçĤ¹
    -0.27
     Benefit
    -0.27
    rike
    -0.26
    POSITIVE LOGITS
    orts
    0.30
    obuf
    0.30
     immediately
    0.29
    argon
    0.28
    ulls
    0.27
     warmly
    0.27
    emons
    0.27
     politely
    0.27
     bombs
    0.27
    .cgi
    0.27
    Activations Density 2.135%

    No Known Activations