© 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. 19688
    Prev
    Next
    INDEX
    Explanations

    Strange characters

    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
    าะ
    -0.28
    ibe
    -0.27
    fully
    -0.27
    él
    -0.27
    ded
    -0.26
    ITU
    -0.26
    DED
    -0.26
    ede
    -0.26
     Murdoch
    -0.26
    azole
    -0.25
    POSITIVE LOGITS
    åĬĪ
    0.27
    大纲
    0.25
    æµ®
    0.25
    æ¢ħ
    0.25
     TableRow
    0.25
    ë°Ģ
    0.25
    éĴī
    0.25
    ursor
    0.25
    æĸĩç«ł
    0.24
    ä»ĬçĶŁ
    0.24
    Activations Density 0.587%

    No Known Activations