© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 18440
    Prev
    Next
    INDEX
    Explanations

    Code and formal documents

    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
    Not in Any Lists

    No Comments

    Negative Logits
    build
    -0.29
    icos
    -0.28
    rice
    -0.27
     cour
    -0.27
    usta
    -0.26
    goog
    -0.26
    woke
    -0.26
     built
    -0.26
    å±ı
    -0.25
     cargo
    -0.25
    POSITIVE LOGITS
    éļIJéļIJ
    0.27
    çį¨ç«ĭ
    0.27
    çĭ¬ç«ĭ
    0.26
    èħĭ
    0.25
    ä¸ŃåIJ«æľī
    0.25
    ä¸Ńå¿ĥ主任
    0.25
     async
    0.25
    å¾Ĺå¾Ī好
    0.24
    éļIJ约
    0.24
    atches
    0.24
    Activations Density 0.103%

    No Known Activations