© 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. 33-TRANSCODER-HP
    7. 53654
    Prev
    Next
    INDEX
    Explanations

    computer code snippets

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ent
    -0.29
    -po
    -0.26
     fat
    -0.25
    å¼µ
    -0.25
    -sl
    -0.25
    fra
    -0.25
    ãĤ¹ãĥĪ
    -0.25
     amp
    -0.25
    害
    -0.24
     fav
    -0.24
    POSITIVE LOGITS
    izu
    0.27
    _LVL
    0.26
    ugg
    0.26
    çĭ¬ç«ĭ
    0.25
    èħ¾
    0.25
    è°ĥåīĤ
    0.25
    _Filter
    0.24
    æ®ĸ
    0.24
     Foot
    0.24
    kaar
    0.24
    Activations Density 0.006%

    No Known Activations