© 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. 19913
    Prev
    Next
    INDEX
    Explanations

    root

    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
    .ua
    -0.28
    ıyla
    -0.24
    æĻ¤
    -0.24
    asmine
    -0.24
     hunted
    -0.24
    .Play
    -0.23
    ç¬ı
    -0.23
    owied
    -0.23
    åŃ£
    -0.23
    åĭĭç«ł
    -0.23
    POSITIVE LOGITS
    æł¹
    1.45
     root
    1.39
    root
    1.27
     roots
    1.27
    Root
    1.19
     Root
    1.16
    roots
    1.11
    _root
    1.10
    	root
    1.02
    -root
    1.02
    Activations Density 6.138%

    No Known Activations