© 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. 19976
    Prev
    Next
    INDEX
    Explanations

    Code and file information

    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
    OrElse
    -0.26
    Root
    -0.26
     masa
    -0.25
     ROOT
    -0.24
    ittle
    -0.24
    roof
    -0.24
    æ··åIJĪ
    -0.24
    æľīæľŁ
    -0.24
    fee
    -0.24
    ideo
    -0.24
    POSITIVE LOGITS
    ä¸Ĭä¸ĭ游
    0.28
    éĢĸ
    0.27
    omes
    0.26
    ivers
    0.26
    edException
    0.25
    æī¾äºĨ
    0.25
    ä¸īåĪĨä¹ĭä¸Ģ
    0.24
     approximately
    0.24
    ä¹ĭå¤ľ
    0.24
    åĽĽæĸ¹
    0.24
    Activations Density 0.002%

    No Known Activations