© 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. 20-TRANSCODER-HP
    7. 108790
    Prev
    Next
    INDEX
    Explanations

    Prepositions

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ä¸įå®ļ
    -0.27
    æĺ¯æ¯Ķè¾ĥ
    -0.26
    ↵  ↵
    -0.25
    èı½
    -0.25
    éĿĴå±±
    -0.25
    è¿Ļæĺ¯
    -0.25
    éļĶéŁ³
    -0.25
    ama
    -0.24
     chances
    -0.24
    ands
    -0.24
    POSITIVE LOGITS
     Futures
    0.28
    appendTo
    0.28
    itur
    0.27
    append
    0.27
    ghi
    0.26
     firm
    0.26
    urry
    0.26
    ÑģÑĤÑĥпил
    0.26
    stered
    0.26
    uhl
    0.25
    Activations Density 25.418%

    No Known Activations