© 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. 19750
    Prev
    Next
    INDEX
    Explanations

    illness, injury

    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
    æŀ¶åŃIJ
    -0.27
    ml
    -0.26
    æĪIJ为ä¸Ģ个
    -0.26
    cplusplus
    -0.25
    åIJĪéĢĤçļĦ
    -0.25
    isable
    -0.25
    KL
    -0.25
     UNDER
    -0.25
    nowledge
    -0.24
    UEST
    -0.24
    POSITIVE LOGITS
     especially
    0.34
    å°¤åħ¶æĺ¯
    0.31
     Especially
    0.30
    å°¤åħ¶
    0.29
     apart
    0.29
     like
    0.28
     spiral
    0.27
    rips
    0.27
    çī¹åĪ«æĺ¯
    0.26
    æĭĻ
    0.26
    Activations Density 1.911%

    No Known Activations